1098 331 FEDERAL SUPPLEMENT, 2d SERIES1098 331 FEDERAL SUPPLEMENT, 2d SERIES allegation of other...

1098 331 FEDERAL SUPPLEMENT, 2d SERIES

allegation of other facts, the dismissal ofthe claims against Philip Morris, Lorillard,R.J. Reynolds, Brown & Williamson, andHill & Knowlton is WITH PREJUDICE.IT IS SO ORDERED.

,

UNITED STATES of America,et al, Plaintiffs,

v.

ORACLE CORPORATION, Defendant.

No. C 04–0807 VRW.

United States District Court,N.D. California.

Sept. 9, 2004.Background: Department of Justice’s An-titrust Division and several states broughtaction under Clayton Act, seeking to en-join proposed merger of two softwaremanufacturers.Holdings: The District Court, Walker,Chief Judge, held that:(1) evidence was insufficient to establish

that relevant product market was lim-ited to ‘‘high function’’ financial man-agement systems (FMS) and humanrelations management (HRM) soft-ware;

(2) evidence established that geographicmarket for the manufacturers’ enter-prise resource planning (ERP) systemsoftware was worldwide, rather thanlimited to the United States; and

(3) evidence was insufficient to show thatmerger would adversely effect compe-tition in localized market, as requiredto support unilateral effects claim.

Judgment accordingly.

1. Monopolies O20(3)To establish a violation of the Clayton

Act’s monopolization provision, plaintiffs

must show that a pending acquisition isreasonably likely to cause anticompetitiveeffects. Clayton Act, § 7, 15 U.S.C.A.§ 18.

2. Monopolies O20(3)

Clayton Act’s monopolization provi-sion does not require proof that a mergeror other acquisition will cause higherprices in the affected market; all that isnecessary is that the merger create anappreciable danger of such consequencesin the future. Clayton Act, § 7, 15U.S.C.A. § 18.


Substantial competitive harm is likelyto result for purposes of Clayton Act’smonopolization provision if a merger cre-ates or enhances ‘‘market power,’’ a termthat has specific meaning in antitrust law.Clayton Act, § 7, 15 U.S.C.A. § 18.


Market definition for purposes ofClayton Act’s monopolization provisionproceeds by determining the marketshares of the firms involved in the pro-posed transaction, the overall concentra-tion level in the industry and the trends inthe level of concentration; a significanttrend toward concentration creates a pre-sumption that the transaction violates toprovision. Clayton Act, § 7, 15 U.S.C.A.§ 18.


Plaintiffs establish a prima facie caseof a violation of the Clayton Act’s monopo-lization provision by showing that themerger would produce a firm controllingan undue percentage share of the relevantmarket, and would result in a significantincrease in the concentration of firms inthat market. Clayton Act, § 7, 15U.S.C.A. § 18.

1099U.S. v. ORACLE CORP.Cite as 331 F.Supp.2d 1098 (N.D.Cal. 2004)

6. Monopolies O20(4)To rebut presumption that post-merg-

er market share of 30 percent or higherviolate’s Clayton Act’s monopolization pro-vision, defendant may show that the mar-ket-share statistics give an inaccurate ac-count of the merger’s probable effects oncompetition in the relevant market. Clay-ton Act, § 7, 15 U.S.C.A. § 18.

7. Monopolies O28(7.1)If the defendant in action under Clay-

ton Act’s monopolization provision success-fully rebuts the presumption of illegality ofa proposed merger, the burden of produc-ing additional evidence of anticompetitiveeffects shifts to plaintiffs, and merges withthe ultimate burden of persuasion, whichremains with the government at all times.Clayton Act, § 7, 15 U.S.C.A. § 18.

8. Monopolies O28(7.1)An application of the burden-shifting

approach under Clayton Act’s monopoliza-tion provision requires the court to deter-mine (1) the line of commerce or productmarket in which to assess the transaction;(2) the section of the country or geograph-ic market in which to assess the transac-tion; and (3) the transaction’s probableeffect on competition in the product andgeographic markets. Clayton Act, § 7, 15U.S.C.A. § 18.

9. Monopolies O20(3)In a structural analysis of proposed

merger, anticompetitive effects are pre-sumed if a plaintiff demonstrates undueconcentration in a well-defined market.Clayton Act, § 7, 15 U.S.C.A. § 18.

10. Monopolies O20(3)To prevail on a differentiated products

unilateral effects claim under ClaytonAct’s monopolization provision, a plaintiffmust prove a relevant market in which themerging parties would have essentially amonopoly or dominant position. ClaytonAct, § 7, 15 U.S.C.A. § 18.

11. Monopolies O20(6.1)Defining the relevant market is criti-

cal in an antitrust case because the legalityof the proposed merger in question almostalways depends upon the market power ofthe parties involved. Clayton Act, § 7, 15U.S.C.A. § 18.

12. Monopolies O20(8)The test of market definition in action

under Clayton Act’s monopolization provi-sion turns on reasonable substitutability;this requires the court to determinewhether or not products have reasonableinterchangeability based upon price, useand qualities. Clayton Act, § 7, 15U.S.C.A. § 18.

13. Monopolies O28(7.5)Evidence in Clayton Act action chal-

lenging proposed merger of two softwaremanufacturers, including customer testi-mony, was insufficient to establish thatrelevant product market was limited to‘‘high function’’ financial management sys-tems (FMS) and human relations man-agement (HRM) software sold by themanufacturers and a competitor; relevantmarket included mid-market vendors, out-sourcing, and ‘‘best of breed’’ solutions.Clayton Act, § 7, 15 U.S.C.A. § 18.

14. Monopolies O20(7)Evidence in Clayton Act action chal-

lenging proposed merger of two softwaremanufacturers established that geographicmarket for the manufacturers’ enterpriseresource planning (ERP) system softwarewas worldwide, rather than limited to theUnited States. Clayton Act, § 7, 15U.S.C.A. § 18.


lenging proposed merger of two softwaremanufacturers was insufficient to showthat a post-merger manufacturer wouldtacitly coordinate by allocating customers


or markets, as required to show thatmerger would have anticompetitive coordi-nated effects. Clayton Act, § 7, 15U.S.C.A. § 18.


lenging proposed merger of two softwaremanufacturers was insufficient to showthat merger would adversely effect compe-tition in localized market for high functionfinancial management systems (FMS) andhuman relations management (HRM) soft-ware, as required to support unilateral ef-fects claim; evidence failed to show thatthere were a significant number of custom-ers who regard the manufacturers as theirfirst and second choices. Clayton Act, § 7,15 U.S.C.A. § 18.

17. Monopolies O20(3)Claimed efficiencies resulting from

proposed merger of two software manufac-turers were insufficient to rebut a showingof anticompetitive effects in Clayton Actaction challenging the merger; estimationsregarding the potential cost-savings to ac-quiring manufacturer were speculative,and its efficiency defense based upon fu-ture innovations was not verified by inter-nal documents or supported by evidenceregarding the functionality or characteris-tics the innovative product would contain,or any evidence regarding its date of avail-ability. Clayton Act, § 7, 15 U.S.C.A.§ 18.

Claude F. Scott, Jr., Conrad JohnSmucker, J. Bruce McDonald, N. ScottSacks, Assistant Chief, Washington, DC,Pamela P. Cole, Phillip R. Malone, PhillipH. Warren, San Francisco, CA, R. HewittPate, Renata B. Hesse, Washington, DC,Chad S. Hummel, Manatt Phelps & Phil-lips LLP, Los Angeles, CA, Kyle DavidAndeer, San Francisco, CA, Greg Abbott,Barry R. McBee, Austin, TX, Edward D.

Burbach, Kim Van Winkle, Mark BernardTobey, Austin, TX, Mark J. Bennett, Hon-olulu, HI, Ellen S. Cooper, Alan MichaelBarr, Gary Honick, J. Joseph Curran, Jr.,John Robert Tennis, Baltimore, MD, Tim-othy E. Moran, Boston, MA, Kristen MarieOlsen, St. Paul, MN, Jay L. Himes, NewYork City, Todd A. Sattler, Bismarck, ND,Arnold B. Feigin, Clare E. Kindall, DarrenP. Cunningham, Steven M. Rutstein, Hart-ford, CT, Beth Ann Finnerty, JenniferLynne Edwards, Mitchell Lee Gentile, Co-lumbus, OH, for Plaintiffs.

Gregory Lindstrom, Latham & WatkinsLLP, J. Thomas Rosch, Latham & Wat-kins LLP, San Francisco, CA, Michele M.Pyle, Latham & Watkins LLP, New YorkCity, Christopher S. Yates, Daniel M.Wall, Latham & Watkins LLP, San Fran-cisco, CA, Joshua N. Holian, Latham &Watkins, San Francisco, CA, for Defen-dant.

FINDINGS OF FACT, CONCLUSIONSOF LAW AND ORDER

THEREON

WALKER, Chief Judge.

The government, acting through the De-partment of Justice, Antitrust Division,and the states of Connecticut, Hawaii, Ma-ryland, Massachusetts, Michigan, Minneso-ta, New York, North Dakota, Ohio andTexas, First Amended Complaint (FAC)(Doc. # 125) ¶ 3 at 5–6, seek to enjoinOracle Corporation from acquiring, direct-ly or indirectly, the whole or any part ofthe stock of PeopleSoft, Inc. Plaintiffs al-lege that the acquisition would violate sec-tion 7 of the Clayton Act, 15 USC § 18.Both companies are publicly traded andheadquartered in this district. Jt. StipFact (Doc. # 218) at 1–2. The court hassubject matter jurisdiction under 15 USC§ 25 and 28 USC §§ 1331, 1337(a) and1345. There is no dispute about the


court’s personal jurisdiction over the de-fendant.

Oracle initiated its tender offer for theshares of PeopleSoft on June 6, 2003. Jt.Stip of Fact (Doc. # 128) at 2; Ex. P2040.Plaintiffs brought suit on February 26,2004. Compl. (Doc. # 1). The case wastried to the court on June 7–10, 14–18, 21–25, 28–30 and July 1, 2004, with closingarguments on July 20, 2004, and furtherevidentiary proceedings on August 13,2004. Based on the evidence presentedand the applicable law, the court concludesthat plaintiffs have failed to carry the bur-den of proof entitling them to relief and,therefore, orders that judgment be en-tered for defendant and against plaintiffs.

INTRODUCTORY FINDINGS:INDUSTRY OVERVIEW

Products at Issue

Of the many types of computer soft-ware, such as operating system software,database software, integration software(sometimes called ‘‘middleware’’ in soft-ware parlance) and utilities software, thiscase involves only one—application soft-ware. And within this type, the presentcase deals with only applications that auto-mate the overall business data processingof business and similar entities; these ap-plications are called ‘‘enterprise applicationsoftware’’ (EAS). Jt. Definitions (Doc.# 332) at 6. There are three main kinds ofEAS. Plaintiffs single out one.

Some EAS programs are mass marketPC-based applications of fairly limited‘‘functionality’’ (meaning capability). Id.(Doc. # 332) at 5. See Daniel E. O’Leary,Enterprise Resource Planning Systems at19 (Cambridge, 2000). Other EAS pro-grams are developed by or for a specificenterprise and its particular needs; mostlarge organizations had such specially de-signed EAS (called ‘‘legacy software’’) pri-or to the advent of the products in suit.Plaintiffs focus their claims on the third,

intermediate category of EAS—enterpriseresource planning (ERP) system software.Jt. Sub. Definitions (Doc. # 332) at 6. ERPis packaged software that integrates mostof an entity’s data across all or most of theentity’s activities. See O’Leary, Enter-prise Resource Planning Systems at 27–38. Oracle and PeopleSoft develop, pro-duce, market and service ERP software.

These copyrighted software programsare licensed (‘‘sold’’ is the term applied tothese license transactions) to end usersalong with a continued right to use licensewhich usually includes maintenance or up-grades of the software. To the customer,the fees to license and maintain ERP soft-ware are generally a small part, 10 to 15percent, of the total cost of the installationand maintenance of an ERP system. Tr.at 133:12–15 (Hatfield); 655:2–4 (Maxwell);1385:6–11 (Gorriz). An ERP installation,because of its complexity, usually requiressubstantial and expensive personnel train-ing, consulting and other services to inte-grate the program into the customer’s pre-existing or ‘‘legacy’’ software. Jt. Sub.Definitions (Doc. # 332) at 6. See alsoO’Leary, Enterprise Resource PlanningSystems at 19. ERP software vendorsoften provide some of those services, butthey are typically also performed and aug-mented by the customer’s own staff, ob-tained from providers other than ERPvendors or both.

Many ERP programs were developed toaddress the needs of particular industries,such as banking and finance, insurance,engineering, construction, healthcare, gov-ernment, legal and so forth (in industrylingo, these are called ‘‘verticals’’). SeeMartin Campbell–Kelly, From AirlineReservations to Sonic the Hedgehog, at169–73 (MIT, 2003). Vertical-specific ERPprograms, although well suited to theneeds of firms engaged in a particularindustry, often are not well suited to the


needs of firms in other verticals. An en-terprise that relies on vertical-specificERP software products, but whose opera-tions embrace more than one vertical facesthe task of integrating the programs. Thelargest and most complex organizationsface particular difficulty. ‘‘[O]nly custom-written software or carefully tailored andintegrated cross-industry packages [can]handle larger firms’ historically idiosyn-cratic accounting systems and diverseoverseas operations.’’ Id. at 172.

ERP programs have been developed tohandle the full range of an enterprise’sactivities; these include human relationsmanagement (HRM), financial manage-ment systems (FMS), customer relationsmanagement (CRM), supply chain man-agement (SCM), Product Life Cycle Man-agement, Business Intelligence (BI),among many others. These are called‘‘pillars.’’ Although ERP encompassesmany pillars, see Ex. D5572, plaintiffs as-sert claims with respect to only two pillars,HRM and FMS. FAC (Doc. # 125) ¶ 23 at12–13.

Within these two pillars, plaintiffs fur-ther limit their claims to only those HRMand FMS products able to meet the needsof large and complex enterprises with‘‘high functional needs.’’ Id. at ¶ 14 at 9.Plaintiffs label HRM and FMS productscapable of meeting these high functionneeds ‘‘high function HRM software’’ and‘‘high function FMS software,’’ respective-ly. Id. ¶ 23(a)-(b) at 12–13. ERP pillarsincapable of meeting these high functionneeds are called ‘‘mid-market’’ software byplaintiffs. Id. ¶ 13 at 9.

‘‘High function software’’ is a termadopted by plaintiffs to describe what theycontend is the separate and distinct line ofcommerce in which they contend competi-tion would be lessened by the proposedacquisition. Id. at ¶ 23 at 13–14. Plain-tiffs apply the term ‘‘high function’’ to bothHRM and FMS. ‘‘High function software,’’

as defined by plaintiffs, has no recognizedmeaning in the industry. See Tr. at 349:7–10 (Bergquist); 2298:6–20 (Elzinga).

Rather, industry participants and soft-ware vendors use the terms ‘‘enterprise’’software, ‘‘up-market’’ software and ‘‘TierOne’’ software to denote ERP that is capa-ble of executing a wide array of businessprocesses at a superior level of perform-ance. See Tr. at 274:24–275:7 (Bergquist);Tr. at 1771:5–1772:1 (Wilmington); Tr. at1554:25–1555:7 (Wolfe); Tr. at 2180:22–2181:5 (Iansiti). Software vendors usethese terms to focus sales and marketinginitiatives. Tr. 2816:6–2818:8 (Knowles)(testifying that SAP divided mid-marketand large enterprise at $1.5 billion basedon SAP’s sales resources and estimatedamount of IT ‘‘spend’’ available from thosecustomers).

Each ERP pillar consists of ‘‘modules’’that automate particular processes or func-tions. HRM and FMS software each con-sists of numerous modules. Exs. P3010,P3011. Tr. at 268:8–269:11, 270:5–271:12(Bergquist). HRM modules include suchfunctions as payroll, benefits, sales incen-tives, time management and many others.Ex. P3010. FMS modules include suchfunctions as general ledger, accounts re-ceivable, accounts payable, asset manage-ment and many others. Ex. P3011.

‘‘Core’’ HRM modules are those specificERP modules that individually or collec-tively automate payroll, employee trackingand benefits administration. Core FMSmodules are those ERP modules that indi-vidually or collectively track general ledg-er, accounts receivable, accounts payableand cash and asset management businessprocesses. Core FMS and HRM modulesare offered by all the ERP vendors thathave HRM and FMS offerings. Ex. P3179(Ciandrini 1/16/04 Dep) at Tr. 256:2–257:10. Large enterprise customers rare-ly, if ever, buy core HRM or FMS modules


in isolation. Tr. at 3461:14–23 (Catz).Customarily, FMS and HRM software arepurchased in bundles with other products.Tr. at 3807:21–3808:1 (Hausman). Seealso Tr. at 3813:12–13 (Hausman). Cus-tomers purchase a cluster of products suchas Oracle’s E–Business Suite that providethe customer with a ‘‘stack’’ of softwareand technology, which may include coreHRM or FMS applications, add-on mo-dules, ‘‘customer-facing’’ business applica-tions such as CRM software, and the infra-structure components (application serversand database) on which the applicationsrun. Tr. at 3461:14–3462:5 (Catz); Tr. at3807:21–3808:1 (Hausman). See, e.g., Exs.P1000–P1322 (Oracle discount requestforms).

ERP vendors, including Oracle and Peo-pleSoft, sell modules individually as well asintegrated suite products. Some ERPvendors sell only one or a few modules.Individual modules are referred to as‘‘point solutions’’ as they address a particu-lar need of the enterprise. ERP vendorsthat sell products for only one or a limitednumber pillars are referred to as pointsolution or ‘‘best of breed’’ providers. Acustomer licensing a particular module be-cause it fits the specific needs of the enter-prise is sometimes said to be seeking abest of breed or point solution. An ERPcustomer that acquires best of breed orpoint solutions faces the task of integrat-ing these solutions with one another andwith the customer’s existing ERP or lega-cy footprint.

Although the production cost of ERPapplications is negligible, vendors bear sig-nificant development and marketing ex-penses and substantial costs of pre- andpostsales support and ongoing mainte-nance and enhancement. ERP vendorsemploy and bear substantial costs of ac-count managers, technical sales forces andpersonnel for user training, product docu-mentation and post-sale support.

Customers at Issue

‘‘Large Complex Enterprises’’ (LCE) isa term adopted by plaintiffs to describethe ERP customers that have ‘‘high func-tion software’’ needs. Based on the testi-mony described hereafter, the court findsthat industry participants and softwarevendors do not typically use this term andit has no widely accepted meaning in theindustry.

While many in the software industrydifferentiate between large customers andmid-market customers, there is no ‘‘brightline’’ test for what is a ‘‘large’’ or ‘‘up-market’’ customer. Tr. 348:23–349:3(Bergquist) (acknowledging ‘‘different par-ties tend to define it differently’’); Tr.2033:1–12 (Iansiti); Ex. P3032 (Henley5/4/04 Dep) at Tr. 98:20–25. Likewise,there is no ‘‘bright line’’ test for what is a‘‘mid-market’’ customer. Tr. at 2820:9–19(Knowles) (SAP executive noting that theseparation between mid-market and largeenterprise customers is ‘‘not an exact sci-ence’’); Ex. D7174 (Pollie 5/26/04 Dep) atTr. 54:14–55:3 (testifying that the meaningof the term mid-market ‘‘varies from, fromeveryone you talk to’’); Ex. P3191 (Block12/16/03 Dep) at Tr. 88:12–21, 94:19–95:3(noting the term mid-market is used inmany different ways by many differentpeople). ERP vendors, analysts, systemsintegrators and others in the industry de-fine the mid-market variously. CompareTr. at 864:19–865:2 (Keating) (noting varia-bility of definitions and that Bearing Pointgenerally refers to mid-market as custom-ers in its General Business Group, which issynonymous with companies having lessthan $2 billion in revenue) with Tr. at1846:17–1847:15 (Wilmington) (PeopleSoftformerly defined mid-market as less than$500 million revenue, but after acquiring JD Edwards, it raised mid-market to in-clude companies with less than $1 billionrevenue).


Prior to Oracle’s tender offer, People-Soft used a proxy of $500 million in reve-nue to distinguish mid-market customersfrom large customers. Tr. at 348:5–18(Bergquist). SAP defines its ‘‘large enter-prise’’ market as companies with morethan $1.5 billion in revenues. Tr. at2819:12–20 (Knowles). Oracle segmentsthe market based on the customers’ reve-nue level or number of employees. Ex.P3070 (Prestipino 5/18/04 Dep) at Tr. at21:5–23:11.

Plaintiffs failed to show ERP vendorsdistinguish mid-market customers fromlarge customers on the amount of moneyspent in an ERP purchase. Yet, as dis-cussed below, this was the basis on whichplaintiffs attempted to quantify the ERPmarket.

Vendors at Issue

Many firms develop, produce, marketand maintain ERP software. Ex. 5543 at8–17. Some ERP software vendors, nota-bly Oracle, PeopleSoft and a German com-pany, SAP AG, developed cross-industryapplications or ‘‘suites’’ of ‘‘generalized in-tegrated software that could be customizedfor virtually any large business,’’ Camp-bell–Kelly, From Airline Reservations toSonic the Hedgehog at 172. It is to theproducts of these three vendors that plain-tiffs direct their allegations. Although notalone in the ERP business, these threefirms have the most comprehensive ERPsoftware offerings.

Oracle. Oracle is headquartered in Red-wood Shores, California. Oracle has over41,000 employees and offices in 80 coun-tries and sells product in over 120 coun-tries. Tr. at 3485:10–12, 3486:16–18(Catz). Oracle’s E–Business suite is a ful-ly integrated suite of more than 70 mo-dules for FMS, internet procurement, BI,SCM, manufacturing, project systems,HRM and sales and service management.Ex. P2209 at xiv. As of December 2002,Oracle had over 5000 customers of its E–

Business Suite, Release lli. Ex. P2208 atORLIT–EDOC–00244117; Ex. P3038.Oracle’s ERP products have enjoyed suc-cess with telecommunications and financialservices customers. Oracle is a major pro-ducer of relational database softwarewhich accounts for a much larger share ofits revenue than its ERP business.

PeopleSoft. PeopleSoft is headquarteredin Pleasanton, California and has 8300 em-ployees. PeopleSoft sells software ‘‘inmost major markets.’’ Ex. 7149 at 7. Ithas offices in Europe, Japan, Asia–Pacific,Latin America and other parts of theworld. Id. PeopleSoft was formed in 1987to develop an HRM product, and it contin-ues to enjoy widespread customer accep-tance of its HRM offerings. PeopleSoftnow sells, in addition to HRM products,FMS, SCM and CRM products and relatedconsulting services. Jt. Stip Fact (Doc.# 218) at 2. In 2003, Peoplesoft generatedabout $1.7 billion in revenue, derived al-most entirely from ERP-related business.PeopleSoft v8 is PeopleSoft’s current inte-grated suite offering. It competes withOracle’s E–Business suite, Release lli.

SAP. SAP AG is headquartered in Wal-dorf, Germany. SAP AG has global opera-tions, including major business operationsin more than a dozen countries and cus-tomers in more than 120 countries aroundthe globe. Tr. at 2805:20–2806:2(Knowles). SAP AG has over 30,000 em-ployees and sells a product called MySAPERP Suite, which includes HRM, FMS,corporate controlling and corporate ser-vices. Tr. at 2811:7–13 (Knowles). SAPAG offers a product called All–in–One,which is ‘‘essentially a scaled-down versionof MySAP ERP with a lot of functionalityturned off.’’ Tr. at 2813:20–2814:2(Knowles). All–in–One is marketed boththrough an indirect channel of resellers tothe $200 million-and-below customer reve-nue segment and by SAP’s direct sales


force. Tr. at 2813:20–2814:2 (Knowles).SAP AG also offers a product called Busi-ness One, which is a ‘‘packaged softwareoffering’’ targeting the $200 million-and-below customer revenue segment and soldthrough an indirect channel of resellers.Tr. at 2813:10–17 (Knowles). SAP has sixsales regions worldwide. SAP America,Inc is responsible for sales in the UnitedStates and Canada. Tr. at 2808:16–19(Knowles). SAP America sells softwaresolutions created by SAP AG. Tr. at2808:8–15, 2806:16–17 (Knowles). In addi-tion to selling software solutions createdby SAP AG, the largest price discountsoffered by SAP America must be approvedby SAP AG. Tr. 2836:22–24 (Knowles).SAP products have won wide acceptance inthe aerospace and petroleum industries.Tr. at 899:9–900:19, 947:10–21 (Keating).

Lawson. Lawson is headquartered inSaint Paul, Minnesota and has 1700 em-ployees. Lawson was founded in the mid–1970s and has 2000 customers, mostly inNorth America and Europe. Lawson of-fers FMS, HRM, procurement products,merchandising products, enterprise per-formance management (EPM), service au-tomation and a unique function called sur-gery instrument management. Tr. at3591:5–10 (Coughlan). In 2003, Lawsongenerated more than $360 million in annu-al revenue. Tr. at 3589:19 (Coughlan).Lawson has tended to do extremely well inthe healthcare and retail verticals. Tr. at3591:1–2 (Coughlan). As Professor JerryHausman testified, and the court will here-after find, although Lawson does not nowcompete in all the industry verticals inwhich Oracle, PeopleSoft and SAP com-pete, Lawson has sufficient resources andcapabilities to reposition to any industryvertical it so chooses. Tr. at 3841:3–13(Hausman).

AMS. AMS is an ERP vendor that wasrecently acquired by CGI, headquarteredin Montreal, Quebec, with offices in North

America, Europe and Asia–Pacific. As anERP vendor, AMS offers FMS, HRM,procurement, tax and revenue software,CRM, CMS, environmental compliancesoftware, performance management andbudgeting and contracting software to gov-ernment entities. See P3034 (Morea5/7/04 Dep) at Tr. 14:19–23. AMS hasbeen successful in its sales to state andfederal governmental agencies, often com-peting head to head with commercial ERPvendors. Tr. at 972:6–15 (Keating) (agree-ing that AMS is a ‘‘viable competitor forlarge and complex federal procurements’’).In fact, only a short time after this actionwas initiated, the United States Depart-ment of Justice chose AMS FMS over theFMS offerings of Oracle, PeopleSoft andSAP. See D7166 (Morea 5/7/04 Dep) at Tr.21:22–22:7.

Microsoft. Microsoft is headquartered inRedmond, Washington, and sells a widerange of software products. In 2001 Mi-crosoft acquired Great Plains Softwareand renamed it Microsoft Great Plains.Microsoft now has a division called Micro-soft Business Solutions (MBS), which wascreated in 2002 when Microsoft GreatPlains acquired the Danish software com-pany Navison. Tr. at 2972:19–2973:9,2973:8–9 (Burgum). MBS has four exist-ing ERP product lines: Navison, GreatPlains, Axapta and Solomon. Tr. at2996:16 (Burgum). Great Plains offersFMS, HRM, E-commerce, retail manage-ment, CRM, analytics and reporting. Seehttp://www.microsoft.com/BusinessSo-luions/GreatPlains/default.aspx. Solomonprovides FMS only. Tr. at 2998:4 (Bur-gum). Navison offers FMS, SCM, CRMand E-commerce. See http://www.micro-soft.com/BusinessSolutions/Navi-son/default.aspx. Finally Axapta offersFMS, HRM, SCM, E-commerce, CRMand analytics. See http://www.micro-soft.com/BusinessSolutions/Axap-ta/default.aspx.


Best of breed vendors. Ninety percentof ERP sales are purchases of software‘‘bundles’’ containing several pillars; rare-ly does a consumer purchase a single pil-lar. Tr. at 3815:10–13 (Hausman). FMSand HRM pillars typically are sold in abundle along with additional kinds of ERP,such as CRM or SCM. Further, the dis-counts that are offered to potential con-sumers are based on the value of the en-tire bundle, not simply based upon thepresence of an HRM or FMS pillar. Tr.at 3813:23–3814:1 (Hausman). According-ly, when Oracle or PeopleSoft offers adiscount on a bundle, it is doing so in orderto ensure that the customer purchases allthe pillars from Oracle or PeopleSoft, rath-er than turn to a best of breed vendor thatspecializes in selling a single kind of pillar.One best of breed vendor, Siebel, sellsindividual pillars of CRM. Testimony sug-gests Siebel is recognized industry-wide asselling high-quality CRM, equal to or bet-ter than the CRM pillars in Tier Onesoftware. Tr. at 3814:15–17 (Hasuman).

Outsourcing. Because of the extensiveamount of training and maintenance in-volved in implementing ERP packagespurchased from ERP vendors, some com-panies have chosen an alternative solu-tion—outsourcing. Outsourcing occurswhen a company hires another firm toperform business functions, often HRMfunctions. Tr. at 2198:15–2198:3 (Elzinga).A company may outsource a single HRMfunction, such as benefits, pensions or pay-roll, or it may choose to outsource itsentire continuum of HRM needs. Tr. at1648:14–22 (Bass). Many firms have out-sourcing capabilities. Some of the out-sourcers discussed at trial include: Accen-ture, Fidelity, ADP, Mellon, Exult, Hewitt,Aon and Convergys. Outsourcing firmsmay process a company’s HR data usingHRM software manufactured by an ERPvendor, such as Oracle, but some outsourc-ing firms use internally created HRM soft-

ware (such as Fidelity using HR Access).Tr. at 3152:18–3153:23 (Sternklar).

In addition to individual vertical success,ERP vendors have tended to enjoy varyingdegrees of success in different geographicregions. SAP, for example, has been moresuccessful at selling ERP to financial insti-tutions in Europe than in North America.Tr. at 996:20–997:15 (Keating).

The FMS and HRM software sold tolarge customers is the same as that sold tomid-market customers. Tr. at 819:8–11(Allen); Tr. at 1787:25–1788:2 (Wilming-ton); Tr. at 3436:24–3437:11 (Catz); Ex.D7166 (Morea 5/17/04 Dep) at Tr. 18:15–19:2(AMS); Ex. P3179 (Ciandrini 1/16/04Dep) at Tr. 235:15–22. All the vendors—-including Oracle, SAP, and PeopleSoft—have a single product ‘‘and that one prod-uct is sold up and down the line’’ to cus-tomers of all sizes. Ex. P3171 (Ellison1/20/04 Dep) at Tr. 148:10–151:15. Whilesome ERP vendors have introduced spe-cial licensing packages of FMS and HRMthat are marketed to smaller customers,the actual software code in the FMS andHRM products sold to both large and mid-market customers is not different. Ex.P3070 (Prestipino 5/18/04 Dep) at Tr.35:19–36:10 (Oracle); Tr. at 3437:5–9(Catz). Oracle has recently launched itsE–Business Suite Special Edition to appealto its smallest customers—those who canuse only 50 seats or less. It contains thesame code as the software sold to thelargest and middle-sized customers, but itarrives pre-configured by the consultingorganization. Tr. at 3436:24–3438:5 (Catz).It contains a subset of the modules foundin Oracle’s E–Business Suite, includingFMS but excluding HRM. Tr. at 3437:5–11(Catz); Ex. P3070 (Prestipino 5/18/04 Dep)at Tr. 25:5–22, 32:19–33:19.

Despite the identity of code in each com-pany’s ERP packaged product, ERP prod-uct offerings are not homogeneous. Whlie


the ERP products offered by Oracle andPeopleSoft and other vendors perform thesame or similar functions, these productsare not uniform in their architecture, scal-ability, functionality or performance char-acteristics. Tr. at 897:23–899:3, 899:9–900:19, 901:6–902:15, 903:6–15, 946:18–20,947:4–9, 992:23–993:7, 993:16–994:2,996:20–997:15 (Keating). The product ofeach vendor possesses certain features orqualities so that none is a perfect substi-tute for any other. As the testimony indi-cated, and the court finds, no vendor iscapable of meeting all of the high functionneeds, as defined by plaintiffs, of all cus-tomers. Tr. at 2085:3–5 (Iansiti).

Furthermore, because each packagedERP product must be customized and con-figured to fit the software footprint of thecustomer, a packaged ERP product may,as fitted to one customer’s informationtechnology footprint, differ significantlyfrom the same packaged ERP product fit-ted to another customer’s footprint. Be-cause of these facts, the court finds theERP products in suit to be differentiatedproducts.

The court also finds that ERP softwareis highly durable and, therefore, regardedby customers as a capital good. Campbelldemo # 5,6,19; see also Tr. at 189:12–18(Hatfield); Tr. at 1107:16–19 (Cichnowicz);Tr. at 1572:14–18 (Wolfe).

Customers almost always purchase acluster of products such as Oracle’s E–Business Suite that provide the customerwith a stack of software and technology,which may include core HRM or FMSapplications, add-on modules, customer-facing business applications such as CRMsoftware and the infrastructure compo-nents (application servers and database)on which the applications run. Tr. at3461:14–3462:5 (Catz); Tr. at 3807:21–3808:1 (Hausman). See, e.g., Exs. P1000–P1322 (Oracle discount request forms).

Plaintiffs’ Claim of ThreatenedInjury to Competition

Plaintiffs allege that the HRM and FMSsold by Oracle, PeopleSoft and SAP arethe only HRM and FMS products that canappropriately be deemed ‘‘high functionHRM and FMS.’’ FAC (Doc. # 125) ¶ 9 at8.

Plaintiffs allege that these ‘‘high func-tion’’ HRM and FMS products have the‘‘scale and flexibility to support thousandsof simultaneous users and many tens ofthousands of simultaneous transactionsand the ability to integrate seamlessly intobundles or ‘suites’ of associated HRM andFMS functions.’’ Id. ¶ 14 at 9. Plaintiffsallege that ‘‘high function’’ HRM and FMSproducts compete in a market that is sepa-rate and distinct from that of all otherERP products, such as SCM, CRM ormid-market HRM and FMS, the latterbeing HRM or FMS products designed fororganizations having less demandingneeds. These mid-market products in-clude Oracle’s E–Business Suite SpecialEdition, SAP’s MySAP and All–in–One,PeopleSoft’s PeopleSoft EnterpriseOneand the products of ERP vendors such asLawson and AMS.

Moreover, plaintiffs allege that this com-petition is geographically confined to theUnited States. Id. at ¶¶ 24, 26 at 13.Within this narrowly defined product andgeographic market, plaintiffs allege thatwith limited and specially explained excep-tions, only Oracle, PeopleSoft and, to alesser degree, SAP’s United States arm,SAP America, are in effective competition.The proposed merger would therefore, inplaintiffs’ view, constrict this highly con-centrated oligopoly to a duopoly of SAPAmerica and a merged Oracle/PeopleSoft.

Oracle, predictably enough, contendsthat plaintiffs’ market definition is legallyand practicably too narrow. Oracle con-tends that (1) ‘‘high function’’ HRM and


FMS software does not exist; ‘‘high func-tion’’ is simply a label created by plaintiffs;(2) there is just one market for all HRMand FMS ERP products; (3) many firmsother than the three identified by plaintiffscompete in the business of developing, pro-ducing, marketing and maintaining HRMand FMS ERP software; (4) this competi-tion plays out in many more products thanthose in the HRM and FMS pillars; (5)price competition comes from sources inaddition to ERP software vendors and in-cludes competition from firms that provideoutsourcing of data processing, the inte-gration layer of the ‘‘software stack’’ andfrom the durability and adaptability of en-terprises’ installed base or legacy systems;(6) the geographic area of competition isworldwide or, at the very least, the UnitedStates and Europe; (7) the knowledgeableand sophisticated customers of ERP soft-ware would impede the exercise of anymarket power by a merged Oracle/People-Soft; and (8) potential entrants are poisedto enter into competition, so that the pro-posed merger will not have an anticompeti-tive effect.

Taking up this dispute, the court firstdiscusses the applicable law and economicprinciples that underlie its decision andthen describes the parties’ contentions andevidence along with the court’s resolutionof the disputed factual issues not previous-ly discussed. This begins with the parties’sharply differing definitions of the productand geographic markets and whetherthere is a level of concentration sufficientto trigger the presumption under UnitedStates v. Philadelphia Nat. Bank, 374 U.S.321, 83 S.Ct. 1715, 10 L.Ed.2d 915 (1963),that the proposed transaction will lead to asubstantial lessening of competition underthe principles set forth in the Departmentof Justice and Federal Trade CommissionHorizontal Merger Guidelines (Apr 2,1992, as revised Apr 8, 1997) (‘‘Guide-lines’’). The court then turns to an effi-

ciency defense offered by Oracle beforesetting forth its conclusions of law.

In brief summary, for the reasons ex-plained at length herein, the court’s find-ings and conclusions are as follows:

1 plaintiffs have not proved that theproduct market they allege, high func-tion HRM and FMS, exists as a sepa-rate and distinct line of commerce;

1 plaintiffs have not proved the geo-graphic market for the products of themerging parties is, as they allege, con-fined to the United States alone;

1 plaintiffs have not proved that a post-merger Oracle would have sufficientmarket shares in the product and geo-graphic markets, properly defined, toapply the burden shifting presump-tions of Philadelphia Nat Bank;

1 plaintiffs have not proved that thepost-merger level of concentration(HHI) in the product and geographicmarkets, properly defined, falls out-side the safe harbor of the HorizontalMerger Guidelines (Guidelines);

1 plaintiffs have not proved that theERP products of numerous other ven-dors, including Lawson, AMS and Mi-crosoft, do not compete with the ERPproducts of Oracle, PeopleSoft andSAP and that these other vendorswould not constrain a small but signifi-cant non-transitory increase in priceby a post-merger Oracle;

1 plaintiffs have not proved that out-sourcing firms, such as Fidelity andADP, would not constrain a small butsignificant non-transitory increase inprice by a post-merger Oracle;

1 plaintiffs have not proved that the abil-ity of systems integrators to adapt,configure and customize competingERP vendors’ products to the needs ofthe group of customers that plaintiffscontend constitute a separate and dis-


tinct product market would not con-strain a small but significant non-tran-sitory increase in price by a post-merger Oracle;

1 plaintiffs have not proved that a post-merger Oracle and SAP would likelyengage in coordinated interaction asthe products of Oracle and SAP arenot homogeneous, but are differentiat-ed products, and that the pricing ofthese products is not standardized ortransparent;

1 plaintiffs have not proved localizedproduct or geographic competition be-tween Oracle and PeopleSoft that willbe lessened as a result of the proposedmerger as the merger would not cre-ate a dominant firm occupying a prod-uct or geographic space in which thereis no serious competition;

1 assuming that localized product orgeographic competition exists betweenOracle and PeopleSoft, plaintiffs havenot proved that SAP, Microsoft andLawson would not be able to reposi-tion themselves in the market so as toconstrain an anticompetitive price in-crease or reduction in output by apost-merger Oracle;

1 plaintiffs have proved that products inthe integration layer of the computersoftware industry and the presence ofincumbent ERP systems would notconstrain anticompetitive conduct onthe part of a post-merger Oracle;

1 Oracle has not proved efficiencies fromthe proposed merger sufficient to re-but any presumption of anticompeti-tive effects; should the court’s princi-pal findings and its conclusion thatplaintiffs have not proved the proposedmerger will likely lead to a substantiallessening of competition not be upheldon appeal, Oracle’s efficiency defenseshould not require further trial courtproceedings.

HORIZONTAL MERGER ANALYSIS

Section 7 of the Clayton Act prohibitsa person ‘‘engaged in commerce or in anyactivity affecting commerce’’ from acquir-ing ‘‘the whole or any part’’ of a business’stock or assets if the effect of the acquisi-tion ‘‘may be substantially to lessen com-petition, or to tend to create a monopoly.’’15 USC § 18. The United States is au-thorized to seek an injunction to block theacquisition, 15 USC § 25, as are privateparties and the several states, Californiav. American Stores Co., 495 U.S. 271, 110S.Ct. 1853, 109 L.Ed.2d 240 (1990); Ha-waii v. Standard Oil Co. of Cal., 405 U.S.251, 258–59, 92 S.Ct. 885, 31 L.Ed.2d 184(1972), and district courts have jurisdictionover such actions. 15 USC § 25; 28 USC§ 1337(a). Plaintiffs have the burden ofproving a violation of section 7 by a pre-ponderance of the evidence.

[1–3] To establish a section 7 violation,plaintiffs must show that a pending acqui-sition is reasonably likely to cause anti-competitive effects. See United States v.Penn–Olin Chem. Co., 378 U.S. 158, 171,84 S.Ct. 1710, 12 L.Ed.2d 775 (1964) (not-ing that a section 7 violation is establishedwhen ‘‘the ‘reasonable likelihood’ of a sub-stantial lessening of competition in the rel-evant market is shown’’); United States v.Marine Bancorp., Inc., 418 U.S. 602, 622–23, 94 S.Ct. 2856, 41 L.Ed.2d 978 (1974);FTC v. H J Heinz Co., 246 F.3d 708, 713,719 (D.C.Cir.2001). ‘‘ ‘Congress used thewords ‘‘may be substantially to lessencompetition’’ (emphasis supplied) to indi-cate that its concern was with probabili-ties, not certainties.’ ’’ Id. at 713 (quotingBrown Shoe Co. v. United States, 370 U.S.294, 323, 82 S.Ct. 1502, 8 L.Ed.2d 510(1962)). ‘‘Section 7 does not require proofthat a merger or other acquisition [will]cause higher prices in the affected market.All that is necessary is that the mergercreate an appreciable danger of such con-


sequences in the future.’’ Hospital Corp.of Am. v. FTC, 807 F.2d 1381, 1389 (7thCir.1986). Substantial competitive harm islikely to result if a merger creates orenhances ‘‘market power,’’ a term that hasspecific meaning in antitrust law. SeeEastman Kodak Co v. Image Tech. Ser-vices Inc., 504 U.S. at 451, 464 (1992);Rebel Oil Co. v. Atlantic Richfield Co., 51F.3d 1421, 1434 (9th Cir.1995).

Market Definition

[4, 5] In determining whether a trans-action will create or enhance market pow-er, courts historically have first defined therelevant product and geographic marketswithin which the competitive effects of thetransaction are to be assessed. This is a‘‘necessary predicate’’ to finding anticom-petitive effects. United States v. E. I. duPont & Co., 353 U.S. 586, 593, 77 S.Ct. 872,1 L.Ed.2d 1057 (1957). Market definitionunder the case law proceeds by determin-ing the market shares of the firms involvedin the proposed transaction, PhiladelphiaNat. Bank, 374 U.S. 321, 83 S.Ct. 1715, 10L.Ed.2d 915, the overall concentration lev-el in the industry and the trends in thelevel of concentration. United States v.Aluminum Co. of Am., 377 U.S. 271, 277–79, 84 S.Ct. 1283, 12 L.Ed.2d 314 (1964);United States v. Von’s Grocery Co., 384U.S. 270, 272–74, 86 S.Ct. 1478, 16 L.Ed.2d555 (1966). A significant trend towardconcentration creates a presumption thatthe transaction violates section 7. UnitedStates v. Baker Hughes Inc., 908 F.2d 981,982–83 (D.C.Cir.1990) (Thomas, J). Seealso United States v. Citizens & SouthernNat. Bank, 422 U.S. 86, 120–22, 95 S.Ct.2099, 45 L.Ed.2d 41 (1975). In otherwords, plaintiffs establish a prima faciecase of a section 7 violation by ‘‘show[ing]that the merger would produce ‘a firmcontrolling an undue percentage share ofthe relevant market, and [would] result [ ]in a significant increase in the concentra-tion of firms in that market.’ ’’ Heinz, 246

F.3d at 715 (quoting Philadelphia Nat.Bank, 374 U.S. at 363, 83 S.Ct. 1715) (al-terations in original). Under PhiladelphiaNat Bank, a post-merger market share of30 percent or higher unquestionably givesrise to the presumption of illegality. 374U.S. at 364, 83 S.Ct. 1715 (‘‘Without at-tempting to specify the smallest marketshare which would still be considered tothreaten undue concentration, we are clearthat 30% presents that threat.’’).

[6, 7] To rebut this presumption, de-fendant may ‘‘show that the market-sharestatistics give an inaccurate account of themerger’s probable effects on competitionin the relevant market.’’ Heinz, 246 F.3dat 715 (internal quotation marks and alter-ations omitted). See also Baker Hughes,908 F.2d at 987; California v. Am. StoresCo., 872 F.2d 837, 842–42 (9th Cir.1989),rev’d on other grounds, 495 U.S. 271, 110S.Ct. 1853, 109 L.Ed.2d 240 (1990); FTCv. Warner Communications, 742 F.2d1156, 1164 (9th Cir.1984); Olin Corp. v.FTC, 986 F.2d 1295, 1305–06 (9th Cir.1993). Arguments related to efficienciesresulting from the merger may also berelevant in opposing plaintiffs’ case. SeeFTC v. Tenet Health Care Corp., 186 F.3d1045, 1054–55 (8th Cir.1999); FTC v. Sta-ples, Inc., 970 F.Supp. 1066, 1088 (D.D.C.1997). ‘‘ ‘If the defendant successfully re-buts the presumption [of illegality], theburden of producing additional evidence ofanticompetitive effects shifts to [plaintiffs],and merges with the ultimate burden ofpersuasion, which remains with the gov-ernment at all times.’ ’’ Heinz, 246 F.3dat 715 (quoting Baker Hughes, 908 F.2d at983) (first alteration in original).

[8] An application of the burden-shift-ing approach requires the court to deter-mine (1) the ‘‘line of commerce’’ or productmarket in which to assess the transaction;(2) the ‘‘section of the country’’ or geo-graphic market in which to assess the


transaction; and (3) the transaction’s prob-able effect on competition in the productand geographic markets. See MarineBancorporation, 418 U.S. at 618–23, 94S.Ct. 2856; FTC v. Harbour Group Invest-ments LP, 1990 WL 198819 at *2 n. 3(D.D.C.1990). See also FTC v. SwedishMatch, 131 F.Supp.2d 151, 156 (D.D.C.2000); FTC v. Cardinal Health, Inc., 12 FSupp 2d 34, 45 (D.D.C.1998); Staples, 970F.Supp. at 1072.

Both the Supreme Court and appellatecourts acknowledge the need to adopt aflexible approach in determining whetheranticompetitive effects are likely to resultfrom a merger. Reflecting their ‘‘generali-ty and adaptability,’’ Appalachian Coals v.United States, 288 U.S. 344, 360, 53 S.Ct.471, 77 L.Ed. 825 (1933), application of theantitrust laws to mergers during the pasthalf-century has been anything but static.Accordingly, determining the existence orthreat of anticompetitive effects has notstopped at calculation of market shares.In Hospital Corp of Am the court upheldthe FTC’s challenge to the acquisition oftwo hospital chains, but noted that ‘‘theeconomic concept of competition, ratherthan any desire to preserve rivals as such,is the lodestar that shall guide the contem-porary application of the antitrust laws,not excluding the Clayton Act.’’ 807 F.2dat 1386. Hence, the court held that it wasappropriate for the FTC to eschew reli-ance solely on market percentages and the‘‘very strict merger decisions of the 1960s.’’Id. at 1386. In addition to market concen-tration, probability of consumer harm inthat case was established by factors suchas legal barriers to new entry, low elastici-ty of consumer demand, inability of con-sumers to move to distant hospitals inemergencies, a history of collusion andcost pressures creating an incentive to col-lude. 807 F.2d at 1388–89.

In United States v. Waste Management,743 F.2d 976 (2d Cir.1984), the court of

appeals reversed a finding of a section 7violation based on market shares and pri-ma facie illegality under Philadelphia NatBank, one made even though there werefew barriers to new entry into the market.The trial court had erroneously ignoredthe Supreme Court’s holding in UnitedStates v. General Dynamics, 415 U.S. 486,94 S.Ct. 1186, 39 L.Ed.2d 530 (1974), that aprima facie case may still be rebutted byproof that the merger will not have anti-competitive effects. A finding of marketshares and consideration of the Philadel-phia Nat Bank presumptions should notend the court’s inquiry.

The trend in these cases away from the‘‘very strict merger decisions of the 1960s,’’Hospital Corp. of Am., 807 F.2d at 1386, isalso reflected in the Guidelines. TheGuidelines view statistical and non-statisti-cal factors as an integrated whole, avoidingthe burden shifting presumptions of thecase law. The Guidelines define marketpower as ‘‘the ability profitably to maintainprices above competitive levels for a signif-icant period of time.’’ Guidelines § 0.1.Five factors are relevant to the finding ofmarket power: (1) whether the mergerwould significantly increase concentrationand would result in a concentrated market,properly defined; (2) whether the mergerraises concerns about potential adversecompetitive effects; (3) whether timelyand likely entry would deter or counteractanticompetitive effects; (4) whether themerger would realize efficiency gains thatcannot otherwise be achieved; and (5)whether either party would likely fail inthe absence of the merger. Guidelines,§ 0.2.

In defining the market, the Guidelinesrely on consumer responses. Starting withthe smallest possible group of competingproducts, the Guidelines then ask ‘‘whether‘a hypothetical monopolist over that groupof products would profitably impose at


least a ‘‘small but significant and nontran-sitory’’ [price] increase [‘‘(SSNIP)’’],’ ’’ gen-erally deemed to be about five percentlasting for the foreseeable future. UnitedStates v. Sungard Data Sys., Inc., 172 FSupp 2d 172, 182 (D.D.C.2001) (quotingGuidelines § 1.11). If a significant num-ber of customers respond to a SSNIP bypurchasing substitute products having ‘‘avery considerable degree of functional in-terchangeability’’ for the monopolist’sproducts, then the SSNIP would not beprofitable. E.I. du Pont, 351 U.S. at 399,76 S.Ct. 994. See Guidelines § 1.11. Ac-cordingly, the product market must be ex-panded to encompass those substituteproducts that constrain the monopolist’spricing. The product market is expandeduntil the hypothetical monopolist couldprofitably impose a SSNIP. Id. § 1.11.Similarly, in defining the geographicalmarket, the Guidelines hypothesize a mo-nopolist’s ability profitably to impose aSSNIP, again deemed to be about fivepercent, in the smallest possible geograph-ic area of competition. Id. § 1.21. If con-sumers respond by buying the productfrom suppliers outside the smallest area,the geographic market boundary must beexpanded. Id.

Once the market has been properly de-fined, the Guidelines set about to identifythe firms competing in the market andthose likely to enter the market within oneyear. Guidelines § 1.32. Following thesesteps, the Guidelines calculate the marketshare of each participant, followed by theHerfindahl–Hirschman Index (HHI) con-centration measurement for the market asa whole. Guidelines § 1.5. The HHI iscalculated by squaring the market share ofeach participant, and summing the result-ing figures. Id. The concentration stan-dards in the Guidelines concern the (1)pre-merger HHI (HHI1), (2) the post-merger HHI (HHI2) and (3) the increasein the HHI resulting from the merger,termed delta HHI (6HHI). See Andrew

I Gavil, William E Kovacic and JonathanB. Baker, Antitrust Law in Perspective;Cases, Concepts and Problems in Competi-tion Policy, 480–84 (Thomson West, 2002).The Guidelines specify safe harbors formergers in already concentrated marketsthat do not increase concentration verymuch. For example if the post-mergerHHI is between 1000 and 1800 (a moder-ately concentrated market) and the 6HHIis no more than 100 points, the merger isunlikely to be presumed illegal. Guide-lines § 1.51. Likewise, if the post-mergerHHI is above 1800 (a highly concentratedmarket) and the 6HHI is no more than 50points, the merger will not be presumedillegal. Id.

Notwithstanding these statistical data,the Guidelines next focus on the likelycompetitive effects of the merger. Guide-lines § 2.0; Baker Hughes, 908 F.2d at 984(‘‘Evidence of market concentration simplyprovides a convenient starting point for abroader inquiry into future competitive-ness * * *.’’). The Guidelines recognizethat anticompetitive effects may arise intwo contexts. First, the Guidelines ad-dress the lessening of competition throughcoordinated interaction between themerged firm and remaining rivals. Guide-lines § 2.1. Second, the Guidelines addressthe anticompetitive effects based on unilat-eral action. Id. § 2.2.

Anticompetitive Effects

Coordinated Effects

In analyzing potential coordinated ef-fects, a court is concerned that the mergermay diminish competition by ‘‘enabling thefirms * * * more likely, more successfully,or more completely to engage in coordina-tion interaction.’’ Guidelines § 2.1. Thisbehavior can be express or tacit (impliedby silence), and the behavior may or maynot be lawful in and of itself. Id. TheGuidelines explicitly recognize that suc-cessful coordinated interaction ‘‘entails


reaching [1] terms of coordination that areprofitable to the firms involved and [2] anability to detect and punish [cheating].’’Id. § 2.1. See also FTC v. Elders Grain,Inc., 868 F.2d 901, 905 (7th Cir.1989);Hospital Corp. of Am., 807 F.2d at 1386–87. Examples of ‘‘terms that are profit-able’’ include common pricing, fixed pricedifferentials, stable market shares and cus-tomer or territorial restrictions. Guide-lines § 2.11

Factors that increase the likelihood ofcoordination include product homogeneity,pricing standardization and pricing trans-parency. Brooke Group Ltd. v. Brown &Williamson Tobacco Corp., 509 U.S. 209,238, 113 S.Ct. 2578, 125 L.Ed.2d 168(1993); Elders Grain, 868 F.2d at 905.Plaintiffs do not contend that any of thoseconditions are presented in the proposedmerger which must, therefore, be analyzedfor unilateral anticompetitive effects.

Unilateral Effects

There is little case law on unilateraleffects merger analysis. Few publisheddecisions have even discussed the issue, atleast using the term ‘‘unilateral effects.’’See, e.g., Swedish Match, 131 F Supp 2dat 168; New York v. Kraft Gen. Foods,Inc., 926 F.Supp. 321, 333–35 (S.D.N.Y.1995); Guidelines § 2.2. But, as the courtdemonstrates below, ‘‘unilateral effects’’ isprimarily a new term to address antitrustissues that courts have in other contextsconsidered for quite some time.

Unilateral effects result from ‘‘the ten-dency of a horizontal merger to lead tohigher prices simply by virtue of the factthat the merger will eliminate direct com-petition between the two merging firms,even if all other firms in the market con-tinue to compete independently.’’ CarlShapiro, Mergers with DifferentiatedProducts, 10 Antitrust 23, 23 (Spring1996). Unilateral effects are thought toarise in primarily two situations, only thesecond of which is alleged in this case.

See Roscoe B Starek III & StephenStockum, What Makes Mergers Anticom-petitive?: ‘‘Unilateral Effects’’ AnalysisUnder the 1992 Merger Guidelines, 63Antitrust LJ 801, 803 (1995); Guidelines§§ 2.21, 2.22; Phillip E Areeda, HerbertHovenkamp & John L. Solow, 4 AntitrustLaw ¶ 910 (Aspen, rev ed 1998) (subdivid-ing unilateral effects theories into fourcategories).

The first situation involves a ‘‘dominantfirm and a ‘fringe’ of competitors produc-ing a homogeneous product.’’ Starek &Stockum, 63 Antitrust LJ at 803. In thissituation, the dominant firm has a substan-tial cost advantage over the fringe compet-itors and, therefore, can restrict output toobtain an above-marginal cost price.

The second situation, and the one hereapplicable, concerns differentiated prod-ucts. Starek & Stockum, 63 Antitrust LJat 803; Guidelines § 2.21. Competition indifferentiated product markets, such asERP products, is often described as ‘‘mo-nopolistic competition.’’ There is a nota-ble and interesting literature on this sub-ject commencing with the path-breakingand independent insights of two notableeconomists. See Edward Chamberlin, TheTheory of Monopolistic Competition (Har-vard, 1933, 1938), Joan Robinson, The The-ory of Imperfect Competition (St Martin’s,1933, 2d ed 1969). The admirably clearexposition found in Paul A. Samuelson &William D. Nordhaus, Economics 187–89(McGraw–Hill, 17th ed 2001) makes appar-ent this nomenclature.

The market demand curve shows thequantity of a good that would be pur-chased in the market at each price, otherthings being equal. Id. at 760. A seller’s‘‘own,’’ or ‘‘residual,’’ demand curve showsthe quantity of the good offered by theseller that would be purchased from theseller at each price, other things beingequal. Under perfect competition, the in-


dividual seller faces a horizontal (each ad-ditional unit brings the same revenue), orperfectly elastic, demand curve becausenothing the seller can do alters demand forthe seller’s product. Id. at 148. The sell-er is a price taker. Because the seller’sdemand curve is horizontal, the seller’s

marginal revenue curve is also horizontaland the seller continues to produce untilits marginal cost is equal to the marketprice or average revenue and profits, aseconomists define them, are zero. See id.fig 8–2 and text at 148–50.

The adjacent figure, borrowed fromSamuelson & Nordhaus, Economics fig 9–4 at 178, illustrates the different picturefacing the monopolist. Its demand curveis not horizontal but reflects the inverserelationship between price and the quanti-ty demanded. Because it is the only sellerof the product, the pure or natural monop-olist faces not the horizontal demand curveof the perfectly competitive firm, but thesloping demand curve of the entire market.

In the graph, the monopolist is able tomaximize profit at the intersection of mar-ginal cost and revenue by reducing outputto 4 and raising the price to $120, whichexceeds marginal cost. The monopolistthus derives a ‘‘monopoly rent’’ equal tothe number of units sold times the differ-ence between the market price (G) and themonopolist’s average cost (F), algebraical-ly, (G—F) x 4. It is this reduction in out-put and elevation of price that has beenthe historic concern of antitrust.


Firms in perfect competition ‘‘producehomogeneous product’’ so that ‘‘price is theonly variable of interest to consumers, andno firm can raise its price above marginalcost without losing its entire marketshare.’’ Jean Tirole, The Theory of Indus-trial Organization at 277 (MIT, 1988).Differentiated products are imperfect sub-stitutes representing as they do differentfeatures or characteristics that appeal var-

iously to different customers. Because noproduct is a perfect substitute of anotherin a differentiated products market, eachseller continues to face a downward slop-ing demand curve. Like a pure monopo-list, the seller of a differentiated product,facing a downward sloping, or less thanperfectly elastic, demand curve, maximizesits profit by pricing above marginal cost.See Samuelson & Nordhaus, Economicsfig 10–3 and text at 188–89.


Like a seller in a perfectively competi-tive market, however, sellers in a ‘‘compet-itive’’ differentiated products market donot obtain monopoly rents. In differenti-ated product markets with few barriers toentry, firms will introduce products thatare increasingly close, although not perfectsubstitutes, for the other products in the

market. The introduction of additionalproducts causes the demand curve facedby each seller to shift downward and left-ward until, at long run equilibrium, thedemand curve intersects the average costcurve of the seller (defined as economistsdefine costs to include a reasonable profit)eliminating the monopolistic rent (ACGB).See id. fig 10–4 and text at 188–89.

Differentiated product markets henceshare some characteristics of both a puremonopoly and perfect competition, in that‘‘prices are above marginal costs but eco-

nomic profits have been driven down tozero.’’ Id. at 189 (describing ‘‘economicprofits’’ as supra-normal profits or monop-oly rents). Firms selling differentiated


products have some ‘‘market power’’ inthat they are able to exert some controlover the prices they obtain although thisdoes not rise to the level of ‘‘monopolypower.’’ See Shapiro, 63 Antitrust LJ 24n. 4 (citing the economic literature).

The Guidelines provide some instructionon the necessary elements of a unilateraleffects claim involving differentiated prod-ucts under section 7.

Substantial unilateral price elevation ina market for differentiated products re-quires [1] that there be a significantshare of sales in the market accountedfor by consumers who regard the prod-ucts of the merging firms as their firstand second choices, and [2] that reposi-tioning of the non-parties’ product linesto replace the localized competition lostthrough the merger be unlikely.

Guidelines § 2.21.Although the Guidelines’ discussion

quoted above may be a helpful start, thefactors described therein are not sufficientto describe a unilateral effects claim.First, the Guidelines’ discussion, at least insection 2.21, emphasizes only the relativecloseness of a buyer’s first and secondchoices. But the relative closeness of thebuyer’s other choices must also be consid-ered in analyzing the potential for priceincreases. The Guidelines later acknowl-edge as much in section 2.212, which rec-ognizes that if a buyer’s other options in-clude ‘‘an equally competitive seller notformerly considered, then the merger isnot likely to lead to a unilateral elevationof prices.’’ Accordingly, a plaintiff mustprove not only that the merging firmsproduce close substitutes but also that oth-er options available to the buyer are sodifferent that the merging firms likely willnot be constrained from acting anticompet-itively.

Second, the Guidelines require only ademonstration of some ‘‘significant shareof sales in the market accounted for by

customers’’ that rank the merging firmsfirst and second. Id. § 2.21. ‘‘Measures ofthe ‘closest substitutes’ or ‘second choices’of inframarginal purchasers of Product Aare only relevant to the degree that infra-marginal and marginal consumers havesimilar preferences. However, essentiallyby definition, marginal and inframarginalconsumers do not share similar prefer-ences.’’ Christopher A. Vellturo, Creatingan Effective Diversion: Evaluating Merg-ers with Differentiated Products, 11 Anti-trust 16, 18 (Spring 1997); Gregory J.Werden & George A. Rozanski, The Appli-cation of Section 7 to Differentiated Prod-ucts Industries: The Market DefinitionDilemma, 8 Antitrust 40, 41 (Summer1994) (‘‘[T]here is no reason why theshares in any delineated market in a dif-ferentiated products industry are indica-tive of the relative importance of eachmerging firm as a direct competitor of theother.’’).

In sum, it appears that four factorsmake up a differentiated products unilater-al effects claim. First, the products con-trolled by the merging firms must be dif-ferentiated. Products are differentiated ifno ‘‘perfect’’ substitutes exist for the prod-ucts controlled by the merging firms. SeeSamuelson & Nordhaus, Economics at187–89; Areeda, Hovenkamp & Solow, 4Antitrust Law ¶ 914d (‘‘By ‘significant’ wemean product differentiation that goes tofairly fundamental differences in productdesign, manufacturing costs, technology, oruse of inputs.’’). Second, the products con-trolled by the merging firms must be closesubstitutes. Products are close substitutesif a substantial number of the customers ofone firm would turn to the other in re-sponse to a price increase. Third, otherproducts must be sufficiently differentfrom the products controlled by the merg-ing firms that a merger would make asmall but significant and non-transitoryprice increase profitable for the merging


firms. Finally, repositioning by the non-merging firms must be unlikely. In otherwords, a plaintiff must demonstrate thatthe non-merging firms are unlikely to in-troduce products sufficiently similar to theproducts controlled by the merging firmsto eliminate any significant market powercreated by the merger. These four factorssubstantially track the analysis in Areeda,Hovenkamp and Solow. Areeda, Hoven-kamp & Solow, 4 Antitrust Law ¶ 914f at68–69.

The essential elements of such a differ-entiated products unilateral effects claimare quite similar to those in ‘‘standard’’antitrust analysis. In standard antitrustanalysis, the court considers both ‘‘demandelasticity’’ and ‘‘supply elasticity’’ in deter-mining whether anticompetitive effects arelikely. Rebel Oil, 51 F.3d at 1436. Inother words, courts determine the degreeto which price increases will cause margin-al buyers to turn to other products ormarginal suppliers to increase output ofthe product. Considerations of demandand supply elasticity also motivate the fac-tors outlined by the court for a differenti-ated products unilateral effects analysis.The factors considering the relative substi-tutability of the products of the mergingand non-merging firms, factors 1 to 3,essentially address demand-side substitut-ability and the repositioning factor, factor4, essentially addresses supply-side substi-tutability.

Antitrust analysis of differentiated prod-uct markets is hardly new. See, e.g., E. I.du Pont, 351 U.S. at 392–93, 76 S.Ct. 994(describing the concepts of monopolisticcompetition and differentiated productmarkets); Areeda, Hovenkamp & Solow, 4Antitrust Law ¶ 914c (suggesting that ear-ly railroad merger cases could be viewedas unilateral effects cases). Indeed, asnoted above, defining a geographic marketinvolves exactly same concept of localized

competition that motivates differentiatedproducts unilateral effects analysis.

Areeda, Hovenkamp and Solow persua-sively contend that ‘‘the appropriate con-clusion [under a unilateral effects analysis]is that the merger has facilitated the emer-gence of a new grouping of sales capable ofbeing classified as a relevant market.’’ Id.¶ 913b. This ‘‘new grouping of sales’’ isone ‘‘in which the merging firms have ei-ther a monopoly or else a dominant share.’’Id ¶ 914f at 69. In an example of twomerging firms, B and C, Areeda, Hoven-kamp and Solow state that ‘‘the mergerdoes not create such a market because acartel of firms B and C would also havebeen able to increase price profitably, indi-cating that B and C were already a rele-vant market.’’ Id. ¶ 914a at 60. But ofcourse, ‘‘before their union, B and C feltone another’s competition, as well as thatof other firms, more significantly than af-ter the merger.’’ Id. Areeda, Hovenkampand Solow also later note that ‘‘the suffi-ciently similar output of other firms mustbe included’’ in the relevant market. Id.¶ 914f at 70.

In a unilateral effects case, a plaintiff isattempting to prove that the merging par-ties could unilaterally increase prices. Ac-cordingly, a plaintiff must demonstratethat the merging parties would enjoy apost-merger monopoly or dominant posi-tion, at least in a ‘‘localized competition’’space.

Unilateral effects analysis shares manysimilarities with standard coordinated ef-fects antitrust analysis. But there are alsonotable differences.

Relevant markets defined in terms of‘‘localized competition’’ may be much nar-rower than relevant markets defined intypical cases in which a dominant positionis required. Judicial experience cautionsagainst the use of qualitative factors todefine narrow markets. This judicial ex-


perience arises, in part, from the rise (andfall) of the ‘‘submarkets’’ doctrine.

In Brown Shoe, the Supreme Court stat-ed that submarkets may constitute rele-vant product markets. ‘‘The outer bound-aries of a product market are determinedby the reasonable interchangeability of useor the cross-elasticity of demand betweenthe product itself and substitutes for it.However, within this broad market, well-defined submarkets may exist which, inthemselves, constitute product markets forantitrust purposes.’’ Brown Shoe, 370U.S. at 325, 82 S.Ct. 1502 (citing E. I. duPont, 353 U.S. at 593–95, 77 S.Ct. 872)(footnote omitted).

Properly construed, Brown Shoe sug-gests merely that the technical definitionof a relevant market in an antitrust casemay be smaller than a layperson wouldnormally consider to be a market. Theuse of the term ‘‘submarket’’ may be usefulin ‘‘overcom[ing] the first blush or initialgut reaction’’ to a relatively narrowly de-fined market. See Staples, 970 F.Supp. at1074 (defining the relevant market as ‘‘thesale of consumable office supplies throughoffice supply superstores’’).

Focusing on ‘‘submarkets’’ may be mis-leading, however, because ‘‘the same proofwhich establishes the existence of a rele-vant product market also shows (or * * *fails to show) the existence of a productsubmarket.’’ H J. Inc. v. InternationalTel. & Tel. Corp., 867 F.2d 1531, 1540 (8thCir.1989). See also Olin, 986 F.2d at 1301.Defining a narrow ‘‘submarket’’ tends torequire a relatively long laundry list offactors, which creates the danger of nar-rowing the market by factors that havelittle economic basis. Courts and commen-tators suggest that the use of the sub-markets doctrine has, in fact, misled courtsinto ‘‘identify[ing] artificially narrowgroupings of sales on the basis of noneco-nomic criteria having little to do with theability to raise price above cost.’’ Areeda,

Hovenkamp & Solow, 4 Antitrust Law¶ 914a at 60. See also Allen–Myland, Inc.v. IBM, 33 F.3d 194, 208 n. 16 (3d Cir.1994); Satellite Television & AssociatedResources v. Continental Cablevision ofVa., Inc., 714 F.2d 351, 355 n. 5 (4thCir.1983).

The similarities between the submarketsdoctrine generally and localized competi-tion in unilateral effects cases are difficultto miss. Indeed, commentators have beenquick to note the potential for ‘‘localizedcompetition’’ analysis to devolve into anunstructured submarket-type analysis.See Areeda, Hovenkamp & Solow, 4 Anti-trust Law ¶ 914a at 60; Starek & Stoc-kum, 63 Antitrust LJ at 814–15 (arguingthat the Guidelines’ focus on localized com-petition should not ‘‘be used as a tool forrehabilitating discredited ‘submarket’ anal-ysis’’).

Furthermore, judicial rejection of mar-kets narrowly defined to a single manufac-turer’s product has been even more pro-nounced than judicial skepticism aboutnarrowly defined submarkets. See, e.g.,E. I. du Pont, 353 U.S. at 592–93, 77 S.Ct.872 (refusing to define a market limited tocellophane); TV Communs. Network, Inc.v. Turner Network Television, Inc., 964F.2d 1022, 1025 (10th Cir.1992) (refusing todefine a market limited to TNT cable pro-vision in the greater Denver area); TownSound & Custom Tops, Inc. v. ChryslerMotors Corp., 959 F.2d 468, 479–80 (3dCir.1992) (en banc) (refusing to define amarket limited to Chrysler products); Gallv. Home Box Office, Inc., 1992 WL 230245at *4 (S.D.N.Y.) (‘‘[T]he natural monopolyevery manufacturer has in its own productsimply cannot serve as the basis for anti-trust liability.’’). Cf. Eastman Kodak, 504U.S. at 481–82, 112 S.Ct. 2072 (upholdingdenial of summary judgment in an install-ed base context).

As emphasized in E. I. du Pont:


[O]ne can theorize that we have monopo-listic competition in every nonstandard-ized commodity with each manufacturerhaving power over the price and produc-tion of his own product. However, thispower that, let us say, automobile orsoft-drink manufacturers have over theirtrademarked products is not the powerthat makes an illegal monopoly. Illegalpower must be appraised in terms of thecompetitive market for the product.

351 U.S. at 393, 76 S.Ct. 994 (footnotesomitted).

Merely demonstrating that the mergingparties’ products are differentiated is notsufficient. Instead, a plaintiff must dem-onstrate product differentiation sufficientto sustain a small but significant and non-transitory price increase.

Additionally, defining markets in termsof ‘‘localized competition’’ may result inmarkets defined so narrowly that one be-gins to question whether the market con-stitutes a ‘‘line of commerce’’ as requiredby section 7. One concern is that the mar-ket is defined so narrowly that it encom-passes an insubstantial amount of com-merce. In Philadelphia Nat Bank, theSupreme Court found a ‘‘workable compro-mise’’ between a geographic market nar-rowly defined in terms of bank offices inthe immediate neighborhood or more ex-pansively defined to include the banksavailable only to large borrowers. 374U.S. at 360–61, 83 S.Ct. 1715. Anotherconcern is that the market is defined sonarrowly it fails to capture the potentialeffects of the merger. For example, itmight be inappropriate to focus on a singlecity in analyzing the effects of a mergerbetween sellers who compete on a muchlarger scale. Cf. Staples, 970 F.Supp. at1073 nn. 5–6 (analyzing the likelihood ofanticompetitive effects in forty-two metro-politan areas).

Even if a narrow market definitionwould be appropriate, it may be more diffi-

cult to identify ‘‘clear breaks in the chainof substitutes’’ sufficient to justify bright-line market boundaries in differentiatedproducts unilateral effects cases. The con-ventional ideal market boundary dividesproducts within the market, which arefreely substitutable with one another, fromproducts outside the market, which arepoor substitutes for the products withinthe market. See United States v. Rock-ford Memorial Corp., 717 F.Supp. 1251,1260 (N.D.Ill.1989) (emphasis added), aff’d,898 F.2d 1278 (7th Cir.1990). In differen-tiated products unilateral effects cases, a‘‘spectrum’’ of product differences, insideand outside the market boundary, is morelikely. In re Super Premium Ice CreamDistribution Antitrust Litig., 691 F.Supp.1262 (N.D.Cal.1988), aff’d. sub. nom.,Haagen–Dazs Co. v. Double RainbowGourmet Ice Creams, Inc., 895 F.2d 1417,1990 WL 12148 (9th Cir.1990) (table). Indiscussing unilateral effects, Shapiro haswritten:

[A]ny attempt to make a sharp distinc-tion between products ‘‘in’’ and ‘‘out’’ ofthe market can be misleading if there isno clear break in the chain of substi-tutes: if products ‘‘in’’ the market arebut distant substitutes for the mergingproducts, their significance may be over-stated by inclusion to the full extent thattheir market share would suggest; andif products ‘‘out’’ of the market havesignificant cross-elasticity with themerging products, their competitive sig-nificance may well be understated bytheir exclusion.

Shapiro, 10 Antitrust at 28. See also Ed-ward Chamberlin, Product Heterogeneityand Public Policy, 40 Am. Econ Rev. (Pa-pers & Procs.) 85, 86–87 (1950).

Additionally, to the extent that clearbreaks are difficult to identify, attempts tocreate defensible market boundaries arelikely to be based on relatively vague prod-


uct characteristics. Product characteris-tics that are too vague do not meet section7’s requirement that the relevant marketbe ‘‘well-defined.’’ See Tenet Health Care,186 F.3d at 1052.

A closer look at product differentiationdemonstrates further difficulties in defin-ing the relevant market in differentiatedproduct unilateral effects cases. Price isone, but only one, of many ways in whichto differentiate a product. A market ofhomogeneous goods can be seen as a mar-ket in which sellers have only one dimen-sion in which to differentiate their product.One expects sellers in such a market to‘‘differentiate’’ their products by loweringthe price until price equals marginal cost.On the other hand, a differentiated prod-uct ‘‘market’’ is a market in which sellerscompete along more dimensions than price.As a result, products competing againstone another in a differentiated productmarket may have widely different prices.That products with widely different pricesmay, in fact, be in the same market com-plicates market definition considerably.

The ‘‘Cellophane fallacy’’ may complicatematters even further. This phenomenontakes its name from an error in the Su-preme Court’s logic E.I. du Pont. In E. I.du Pont, the plaintiff was the primarymanufacturer of cellophane. The SupremeCourt held that the relevant market in-cluded ‘‘all flexible wrappings’’ becausecross-price elasticities of demand indicatedthat an increase in the price currentlycharged for cellophane would cause a sig-nificant number of purchasers to turn toother flexible wrapping products.

The error in the logic of E. I. du Pont isthat ‘‘ ‘[t]he existence of significant substi-tution in the event of further price increas-es or even at the current price does nottell us whether the defendant already ex-ercises significant market power.’ ’’ East-man Kodak, 504 U.S. at 471, 112 S.Ct.2072 (quoting Phillip Areeda & Louis Ka-

plow, Antitrust Analysis ¶ 340(b) (Aspen,4th ed 1988)). Stated slightly differently,because a monopolist exercises marketpower by increasing price until the cross-price elasticity of demand is so high that afurther price increase would be unprofit-able, a high cross-price elasticity of de-mand at current prices, by itself, does notdemonstrate that the seller lacks marketpower.

The implications of the Cellophane falla-cy on market definition in differentiatedproduct market cases would seem to sug-gest caution. Courts should be wary ofdefining markets so broadly that a seller’sexisting market power is missed. On theother hand, in differentiated product mar-kets, some measure of market power isinherent and an unduly narrow productmarket definition proves too much. Inmerger analysis, the court is concernedprimarily with determining whether themerger would enhance market power, notwhether market power currently exists.

In sum, defining the relevant market indifferentiated product markets is likely tobe a difficult task due to the many non-price dimensions in which sellers in suchmarkets compete. Further, it may be dif-ficult to determine currently existing mar-ket power and separate this from en-hanced market power due to the merger.

The inability clearly to define a marketsuggests that strong presumptions basedon mere market concentration may be ill-advised in differentiated products unilater-al effects cases. As noted by Starek andStockum, ‘‘it is generally misleading tosuggest that a firm ‘‘controls’’ a certainmarket share in the absence of an analysisbeyond market concentration.’’ Starek &Stockum, 63 Antitrust LJ at 804. See alsoJerry A. Hausman & Gregory K. Leonard,Economic Analysis of DifferentiatedProducts Mergers Using Real World Data,5 Geo. Mason L. Rev. 321, 337–39 (1997).


Such a concern applies with equal force todifferentiated products unilateral effectsclaims. Furthermore, in differentiatedproducts unilateral effects cases, the merg-ing parties’ combined market shares rela-tive to competitors may be less relevantthan the size of their market shares indetermining whether anticompetitive ef-fects are likely. See Gregory J. Werden &Luke M. Froeb, The Effects of Mergers inDifferentiated Products Industries: LogitDemand and Merger Policy, 10 J. L.Econ. & Org. 407, 413 (1994).

Accordingly, a strong presumption ofanticompetitive effects based on marketconcentration is especially problematic in adifferentiated products unilateral effectscontext.

Despite the problems with qualitativeanalyses, modern econometric methodshold promise in analyzing differentiatedproducts unilateral effects cases. Mergersimulation models may allow more preciseestimations of likely competitive effectsand eliminate the need to, or lessen theimpact of, the arbitrariness inherent indefining the relevant market. For exam-ple, some merger simulation methods com-pensate for potential errors in market defi-nition. A model advanced by Werden andFroeb uses a set of ‘‘inside goods’’ and aset of ‘‘outside goods.’’ Id. at 410. Themodel contains a parameter, beta, thatcontrols for the substitutability among theinside goods and another parameter, epsi-lon, that controls for the substitutabilitybetween the inside and outside goods. Id.To the extent the set of goods consideredas ‘‘inside goods’’ is defined narrowly, epsi-lon increases. Id. at 424–25. The in-crease in epsilon increases the predictedamount of substitution to outside goods.Accordingly, error in defining the productmarket too narrowly will be offset, at leastto some extent, by the increase in epsilon.

In sum, differentiated products unilater-al effects analysis shares many similarities

to ‘‘standard’’ antitrust analysis. The pri-mary differences are that the relevantmarket is likely to be smaller and moredifficult to define and that quantitativeanalyses may be robust.

[9] In analyzing antitrust claims,courts have considered both ‘‘circumstan-tial’’ and ‘‘direct’’ evidence of anticompeti-tive effects. See Rebel Oil, 51 F.3d at1434. Even though ‘‘direct’’ evidence ofthe potential for anticompetitive harmfrom a merger is not literally available,merger analyses range from highly quali-tative (‘‘circumstantial’’) to highly quantita-tive (‘‘direct’’), depending on the data avail-able for a particular market. Qualitativeanalyses of antitrust claims are most oftenstructural. In a structural analysis, anti-competitive effects are presumed if a plain-tiff demonstrates undue concentration in awell-defined market. See PhiladelphiaNat. Bank, 374 U.S. at 363, 83 S.Ct. 1715;Baker Hughes, 908 F.2d at 982. A rele-vant market may be defined by referenceto Brown Shoe ’s ‘‘practical indicia.’’ 370U.S. at 325, 82 S.Ct. 1502. Once the rele-vant market is defined, market shares arecalculated and inferences are drawn fromthe degree of concentration.

The Guidelines adopt a structural ap-proach for addressing unilateral effectsclaims that closely mirrors traditionalstructural analysis. See Guidelines§ 2.211. The biggest weakness in theGuidelines’ approach appears to be itsstrong reliance on particular market shareconcentrations. Under the Guidelines,anticompetitive effects are presumed‘‘[w]here market concentration data falloutside the safeharbor regions of Section1.5, the merging firms have a combinedmarket share of at least thirty-five per-cent, and where data on product attributesand relative product appeal show that asignificant share of purchasers of onemerging firm’s product regard the other as


their second choice,’’ unless ‘‘rival sellerslikely would replace any localized competi-tion lost through the merger by reposition-ing their product lines.’’ Id. at §§ 2.211,2.212.

[10] A presumption of anticompetitiveeffects from a combined share of 35% in adifferentiated products market is unwar-ranted. Indeed, the opposite is likely true.To prevail on a differentiated products uni-lateral effects claim, a plaintiff must provea relevant market in which the mergingparties would have essentially a monopolyor dominant position. In Rebel Oil, theNinth Circuit noted that a market share of30% is ‘‘presumptively insufficient to es-tablish the power to control price.’’ 51F.3d at 1438.

Market definitions, statistical presump-tions and likelihood of unilateral anticom-petitive effects are all issues on which theparties contended vigorously and present-ed much evidence. To these, the courtnow turns.

CONTENTIONS, EVIDENCEAND FINDINGS

[11] ‘‘Defining the relevant market iscritical in an antitrust case because thelegality of the proposed merger[ ] in ques-tion almost always depends upon the mar-ket power of the parties involved.’’ Cardi-nal Health, 12 F.Supp.2d at 45. Yet theprecise characteristics that plaintiffs haveused to describe the line of commerce al-legedly affected by the proposed transac-tion changed throughout the course of thislitigation. And the evidence of marketshares presented to enable the court toapply the Philadelphia Nat Bank pre-sumptions or make the HHI calculations ofthe Guidelines is, given the mountain ofevidence plaintiffs presented, startlingsparse.

Plaintiffs’ Proposed ProductMarket Definition

Plaintiffs offer a product market of highfunction HRM and FMS and a geographicmarket of the United States.

Four elements constitute plaintiffs’ defi-nition of high function HRM software asalleged in the FAC: ‘‘[1] Human ResourceManagement (HRM) [2] software and ac-companying services [3] that can be inte-grated into suites of associated functionsfrom a single vendor [4] with performancecharacteristics that meet the demands ofmultifaceted organizations with high-levelfunctional needs.’’ FAC (Doc. # 125)¶ 23(a) at 12.

Likewise, four elements constitute plain-tiffs’ definition of high function FMS soft-ware as alleged in the FAC: ‘‘[1] FinancialManagement Services (FMS) [2] softwareand accompanying services [3] that can beintegrated into suites of associated func-tions from a single vendor [4] with per-formance characteristics that meet the de-mands of multifaceted organizations withhigh-level functional needs.’’ Id. ¶ 23(b) at12–13.

The FAC also notes certain performancecharacteristics of high function software:

Customers with high-level functionalneeds (‘‘enterprise customers’’) requireproducts that can support their ongoingbusiness processes and reporting re-quirements that may stretch across mul-tiple jurisdictions (often requiring sup-port for foreign languages and reportingrequirements), multiple legal entities ordivisions within the organization andmultiple lines of business. These prod-ucts must have the scale and flexibilityto support thousands of simultaneoususers and many tens of thousands ofsimultaneous transactions, and the abili-ty to integrate seamlessly into bundlesor ‘‘suites’’ of associated HRM and FMSfunctions. Most importantly, these


products must have the flexibilitythrough configuration options or othermeans to be matched to the administra-tive and reporting processes of eachunique customer.

Id. ¶ 14 at 9.Plaintiffs clarified their allegations at

the request of the court during the trial bysubmitting a statement of definitions, someof which were joined by defendant. Jt.Sub. Definitions (Doc. # 332). In thesedefinitions, plaintiffs omitted ‘‘and accom-panying services’’ from the second elementalleged in the FAC. Plaintiffs also relegat-ed the FAC’s third element regarding inte-gration to a mere sub-element of the per-formance characteristics described in theFAC’s fourth element. Finally, plaintiffsdescribe four ‘‘performance capabilities.’’Products in the market are (1) ‘‘highly’’configurable, (2) ‘‘seamlessly’’ integratedsoftware products that support (3) ‘‘multi-ple’’ languages, currencies and legal re-gimes with (4) ‘‘virtually unlimited’’ scala-bility. See id. at 2–4 & n. 2.

This definition shifted somewhat inplaintiffs’ proposed findings of fact andconclusions of law. Plaintiffs clarified thedefinition to include ‘‘licensing and mainte-nance’’ rather than ‘‘licensing and accom-panying services,’’ as alleged in the FAC’ssecond element. See Pls. Prop FF (Doc.# 356) at ¶¶ 3.1.1—3.1.2. Plaintiffs alsoadded an element to the formal definition,claiming that high function software ‘‘pro-vid[es] robust functionality that allows or-ganizations to go beyond the basics.’’ Id.at ¶ 3.1.3.4.

Even though not stated as part of theformal definition of high function software,plaintiffs scatter throughout their pro-posed findings of fact other characteristicsof ERP software in an apparent attemptfurther to narrow the relevant market.

First, plaintiffs point to the claimedstrength of high function software in

‘‘core’’ applications. See, e.g., id. at¶ 3.6.2.1.

Second, plaintiffs emphasize that highfunction customers purchase high functionsoftware. Id. (Doc. # 356).

Third, plaintiffs emphasize the brandvalue of the software vendor. Factors thatpromote vendor brand value include previ-ous experience in a particular industry,research and development spending andlocal sales forces. See, e.g., id. (Doc.# 356) at ¶¶ 3.2.4.3—3.2.4.5.

Fourth, plaintiffs note the incumbent ad-vantage software vendors have in compet-ing for further sales with a customer whohas that vendor’s product as part of itsexisting footprint. See, e.g., id. at¶ 7.3.2.1.18 (pointing to testimony that‘‘Bearing Point has identified more than1,200 companies that now have an OracleFinancials and PeopleSoft HR footprint’’).

Fifth, plaintiffs emphasize the allegedstrength of Oracle and PeopleSoft in cer-tain industry verticals, such as insurance.See, e.g., id. at ¶¶ 7.2.3.10—7.2.3.13.

Sixth, plaintiffs describe high-functionsoftware as being ‘‘Able to AccommodateRapid Growth, Acquisitions and Reorgani-zations.’’ Id. at ¶ 2.2.5.

Seventh, plaintiffs define high functionsoftware as allowing users to consolidatedata across multiple organizations whilestill allowing the user to drill down to theoriginal data. Id. at ¶ 2.2.6.

In their post-trial brief too, plaintiffsadjusted their proposed product marketdefinition. They eliminated the ‘‘robustfunctionality’’ factor and incorporated twoof the factors scattered throughout theirproposed findings of fact into the moreformal definition of high function software.The newly incorporated factors are thathigh function software must accommodaterapid growth and complicated businessstructures.


At closing argument, plaintiffs dis-claimed reliance on high function soft-ware’s claimed strength in ‘‘core’’ function-ality in defining the relevant market andaccused Oracle of creating confusion ‘‘bylimiting the relevant market to basic ‘core’functionality.’’ Pl. Post Brief (Doc. # 366)at 10 n. 17.

Added together, plaintiffs propose avery restricted product market definition:HRM and FMS integrated suites sold tolarge complex enterprises (‘‘high functionFMS and HRM market’’). See id (Doc.# 366) at 8. Plaintiffs have defined theasserted relevant product market using alarge number of factors. In sum, the com-petition between Oracle and PeopleSoftthat plaintiffs claim will be impaired bearsthe following characteristics:

Product characteristics:1 Software licensing and maintenance;1 HRM and FMS (as separate markets);Customer characteristics:1 High function needs;1 Oracle or PeopleSoft are major ven-

dors in their software footprint;Performance characteristics:1 Scalable;1 Highly configurable;1 Seamlessly integratable;1 Able to accommodate rapid growth,

acquisitions and reorganizations;1 Able to reflect actual units of business;

and1 Able to adapt to industry specific re-

quirements.Plaintiffs contend that this product marketdoes not include mid-market vendors, best-of-breed solutions, incumbent solutions oroutsourcing. Id. (Doc. # 366) at 14–19.

Plaintiffs’ Evidence of a High FunctionHRM & FMS Market

In support of their proposed productmarket definition and theory of anticom-petitive effects, plaintiffs presented at trial

or through deposition ten customer wit-nesses, five industry witnesses, two sys-tems integration witnesses, three expertwitnesses, a few others who appear mostlyto have been presented to fill a gap or twoin the evidence or, because every trialseems to need some, for spice (e.g., theEllison and Phillips videotape depositiontestimony) and a plethora of exhibits, someof these also for spice (e.g., Ex. P2290).The court will not attempt to recount oreven summarize the entire evidentiary rec-ord. Given the quantity of evidence, thatwould be unduly time-consuming and isunnecessary. It suffices to note that thelaboring oar of the plaintiffs’ case waspulled by the customer witnesses (whomplaintiffs’ counsel described as theirstrongest witnesses), by some of the sys-tems integrator and industry witnessesand by the experts.

Customer Witnesses

Michael Gorriz, Vice President of Infor-mation Technology Business at Daimler-Chrysler (Daimler), testified about hiscompany’s large and complex needs re-garding HRM software. Tr. at 1368 (Gor-riz). Daimler has about 365,000 employeesworldwide in about 100 manufacturing fa-cilities. Tr. at 1368:6–13 (Gorriz). Since1996, Daimler has used SAP as its finan-cial management software. Tr. at 1370:4–10 (Gorriz). Daimler requires highly func-tional HRM to accommodate its largenumber of employees and to comply withthe differing labor laws and union agree-ments in different countries. Tr. at1371:9–12 (Gorriz). For its HRM needs,Daimler currently uses PeopleSoft. Daim-ler chose PeopleSoft based upon its repu-tation and the fact that companies of com-parable size to Daimler have had successwith PeopleSoft HRM. Tr. at 1375:13–21(Gorriz). But when Daimler was firstsearching for an HRM vendor in 1996,Gorriz stated that ‘‘only SAP, PeopleSoft


or Oracle could serve [Daimler’s] needs forthe HR management.’’ Tr. at 1376:9–11(Gorriz). Gorriz stated that Daimler con-sidered no other vendors. Tr. at 3716:18–19 (Gorriz). Daimler’s legacy system was‘‘too old’’ for the company seriously toconsider upgrading. Tr. at 1376:24 (Gor-riz). Daimler did not consider outsourcingto be an option because Daimler’s HRMrequirements were, Gorriz testified, ‘‘toocomplex.’’ Tr. at 1377:24–25 (Gorriz).Further, if Oracle, SAP or PeopleSoftwere to increase their price for HRM by10 percent, Gorriz stated that Daimler‘‘would not consider any offer’’ from anyother vendors. Tr. at 1381:16 (Gorriz).

Bob Bullock, Senior Vice President andChief Information Officer of CH2M Hill,testified about the ERP needs of that civiland environmental engineering firm.CH2M Hill has 14,000 employees, 200worldwide offices and over $2 billion inannual revenue. CH2M Hill has used Or-acle FMS since 1993, but in 2002 the com-pany decided to replace its legacy HRMsoftware. Bullock stated that throughconsultation with the Gartner Group,CH2M Hill was given a list of HRM ven-dors. CH2M Hill did not seriously consid-er SAP, as it ‘‘was a very complex prod-uct’’ and had a ‘‘reputation for being acostly product.’’ Tr. at 207:19–20 (Bull-ock). In Bullock’s opinion, there wereonly two candidates, Oracle and People-Soft. Id at 208:7–8 (Bullock). CH2M Hillnever considered outsourcing, Lawson orremaining on its legacy system. Tr. at210:8, 211:12, 216:8–9 (Bullock). Oracleand PeopleSoft both offered initial bidsbetween $1.5 and 41.6 million. Bullockstated that if this price had been 10 per-cent higher, CH2M Hill would not walkedaway from the deal with Oracle or People-Soft. Tr. at 218–19 (Bullock).

Curtis Wolfe, CIO for the State of NorthDakota, testified about the state’s processof picking an ERP vendor. Tr. at 1532

(Wolfe). North Dakota has approximately10,000 full and part-time employees, 58state agencies and a budget of $5 billion.Tr. at 1533 (Wolfe). In 2002, the statedecided to buy a full ERP program thatincluded FMS and HRM. Tr. at 1534:10–16(Wolfe). North Dakota had a unique needin that it required that its ERP serve thestate’s higher education facilities as well.Id. North Dakota had six vendors submitproposals: Oracle, PeopleSoft, SAP, SCT,Jenzabar (a partner of Lawson) and Micro-soft’s Great Plains. Tr. at 1543:21–22(Wolfe). The state eliminated SAP, GreatPlains and Jenzabar almost immediately.SAP was too expensive, while Jenzabarand Great Plains did not have the requiredfunctionality. Tr. at 1545–46 (Wolfe).SCT did not make the final round; whileSCT met the functionality for the highereducation area, it could not do so withstate agency needs. Tr. at 1551:1–4(Wolfe). Oracle and PeopleSoft were inhead to head competition and Wolfe testi-fied that he believes that this caused thestate to get a $6 to $8 million lower finalbid from each vendor. Tr. at 1561:10–11(Wolfe). If these final offers had been 10percent higher, Wolfe stated that NorthDakota would not have turned to Lawson,Microsoft, SCT, outsourcing or writing itsown software. Tr. at 1569–1570 (Wolfe).

Kenneth Johnsen, Chief of Technologyfor Pepsi Americas, testified as to his con-cerns about the Oracle/PeopleSoft merger.Pepsi Americas is the second largest bott-ler of Pepsi-brand soft drinks within thePepsi system and the third largest bottlerworldwide. Tr. at 1723:25–1724:1 (John-sen). Pepsi Americas has over 15,000 em-ployees and annual revenues of about $3.2billion. Tr. at 1724:5–10 (Johnsen). PepsiAmericas uses PeopleSoft ERP in itsNorth America operations and SAP ERPin its European operations. Tr. at1727:13–14 (Johnsen). Johnsen testifiedthat he has ‘‘a concern’’ about the impact


of this merger on the long-term effective-ness of the PeopleSoft ERP. Tr. at 1734:23(Johnsen). Johnsen is concerned that apost-merger Oracle, while agreeing tomaintain the PeopleSoft ERP, will not pro-vide enhancements to the functionality ofthe software (i.e., upgrades). Tr. at 1737:1–9 (Johnsen). To Johnsen this leavesPepsi Americas with two options: con-stantly upgrade with point solutions (nothis desired choice) or buy ERP from a newvendor. When asked what vendors hecould turn to meet his ERP needs, John-sen claims there are no options outside ofOracle, PeopleSoft and SAP. Tr. at 1739:14(Johnsen).

Scott Wesson, Senior Vice President andChief Information Officer of AIMCO, dis-cussed the company’s choices for FMS andHRM software. Tr. at 1126 (Wesson).AIMCO is the largest owner and operatorof apartment buildings in the UnitedStates. Tr. at 1127:7–8 (Wesson). Thecompany owns approximately 2000 com-plexes in 47 states and the District ofColumbia. AIMCO has over 6,500 employ-ees and an annual revenue of about $1.5billion. Tr. at 1127:9–24 (Wesson). Forits FMS, AIMCO uses PeopleSoft’s finan-cial suite. For its HR payroll systems,AIMCO currently uses Lawson. Tr. at1129:8,21 (Wesson). In 2002, AIMCO be-gan to reevaluate its HRM options and ithired Towers Perrin consult in this pro-cess. Towers Perrin told AIMCO thatonly three vendors could meet AIMCO’sHRM needs: PeopleSoft, Oracle and SAP.Tr. at 1132:7–8 (Wolfe). (There was noobjection to the question that elicited thisresponse). Wesson stated that AIMCOdecided not to upgrade to the latest ver-sion of Lawson because it would have costAIMCO ‘‘about the same * * * as it wouldto go with a new system’’ and also, Lawson‘‘[was] lacking some key features’’ thatAIMCO was looking for. Tr. at 1133:5–11(Wolfe). AIMCO was deciding betweenOracle and PeopleSoft when Oracle first

made its tender offer to PeopleSoft. Tr.at 1143: 9–10 (Wolfe). Wesson stated thatbecause of this proposed merger, he be-lieves PeopleSoft gave him a ‘‘very gooddeal’’ on the HRM. Tr. at 1144:17 (Wolfe).Wesson testified that Oracle agreed tomatch any price offered by PeopleSoft.Tr. at 1145:5 (Wolfe). Wesson said AIM-CO ultimately chose PeopleSoft becausePeopleSoft had guaranteed to pay AIMCOthree times the contract price should therebe a ‘‘change of ownership’’ at PeopleSoft.Tr. at 1146:14, 1147:6–16 (Wolfe). AIMCOis expecting to implement the PeopleSoftsystem in late 2004 or early 2005. Tr. at1148:10 (Wolfe). Moreover, Wesson stat-ed, AIMCO does not consider outsourcingto be a viable option because it is not quickto respond to ‘‘last minute changes,’’ suchas new benefits programs. Tr. at 1150:10(Wolfe). Best of breed solutions are tooexpensive for AIMCO to consider. Tr. at1150:22–24 (Wolfe).

Richard Cichanowicz, Vice President ofSystems Integration of Nextel, testifiedabout the wireless services company’sERP needs. Nextel has 13 million sub-scribers, over $8 billion in annual revenue17,000 [transcript incorrect] employees.See Tr. at 1052:25–1053:3 (Cichanowicz).Before 2002, Nextel had been using Peo-pleSoft HRM, Oracle FMS and AribaSCM. Tr. at 1058:9–11 (Cichanowicz). In2002, however, Nextel determined that us-ing one integrated solution would providemore operational efficiency. Tr. at 1061:7–9 (Cichanowicz). Nextel received advicefrom six consulting firms, which informedNextel that Oracle, SAP and PeopleSoftcould meet those software needs. Tr. at1066:13–19 (Cichanowicz). Nextel thensent RFPs to Oracle and Peoplesoft. Tr.at 1067:25–1068:3 (Cichanowicz). Nexteldid not seriously consider SAP because itwas already using Oracle for FMS andPeopleSoft for HRM and believed thatconversion costs and risks for those two


vendors would be lower. Tr. at 1068:4–17(Cichanowicz). Nextel ultimately chosePeoplesoft, based on its scoring of vendorcriteria such as functionality, ease of inte-gration, scalability, audits, costs and rela-tionship confidence. See Tr. at 1071:20–1072–22 (Cichanowicz). Even after it hadchosen PeopleSoft, however, Nextel contin-ued to negotiate with Oracle for leveragepurposes until the signing of the December2002 contract with PeopleSoft. Tr. at1073:11–20 (Cichanowicz). Cichanowiczstated that if the price of the Oracle orPeopleSoft licenses had been 10 percenthigher, Nextel would not have considereda best of breed approach, writing or build-ing its own ERP software, outsourcing,staying with its previous system or usingSAP or any other United States vendor.Tr. at 1077:16–1080:25 (Cichanowicz).

Mary Elizabeth Glover, Vice Presidentof Information Technology at GreyhoundLines, testified about her company’s forayinto the market for HRM software.Greyhound is in the bus transportationbusiness in both the United States andCanada. The company employees some16,000 people and has annual revenues ofaround $1.2 billion. Tr. at 1459–1460(Glover). For its FMS, Greyhound usesOracle in the United States and J D Ed-wards in Canada. Tr. at 1464:11–21 (Glo-ver). For its HRM, Greyhound uses aproduct called HR1 in the United Statesand HR2000 in Canada. The companyoutsources its payroll to ADP. Tr. at1465:11 (Glover). Glover stated that theHR incumbent systems are ‘‘very old’’ andno longer meet the needs of the company.Tr. at 1466:21–25 (Glover). Further, shetestified that outsourcing is too expensivefor Greyhound. Tr. at 1467:12–15 (Glo-ver). For these reasons, in 2001, Grey-hound began a potential procurement pro-cess for new HRM software. Tr. at1468:17–18 (Glover). The company hiredCDG & Associates to match Greyhoundwith potential vendors who met their

HRM needs. The firm narrowed the se-lection down to only four vendors: Oracle,PeopleSoft, Lawson and Ultimate Soft-ware. Tr. at 1470:11 (Glover). Grey-hound never considered SAP because theconsulting firm believed they were toocostly. Tr. at 1470:16 (Glover). UltimateSoftware was eliminated soon thereafterbecause of lack of functionality. Tr. at1470:24–25 (Glover). Greyhound eliminat-ed PeopleSoft as being too costly. Be-tween Oracle and Lawson, Greyhoundfound that Oracle had more functionality;therefore, Lawson was eliminated. Butbefore Greyhound made a final choice,Glover stated that the company decided togive PeopleSoft a second look. Upon re-examination, Greyhound determined thatboth Oracle and PeopleSoft could meetthe company’s needs, with the companypreferring PeopleSoft over Oracle. Tr. at1483:6–9 (Glover). Unfortunately, theevents of September 11, 2001, a new CEOand a decrease in profits caused Grey-hound to lose the funds necessary to pur-chase the software. Tr. at 1490:6–11(Glover). But Glover stated that shouldGreyhound ever decide to purchase HRMsoftware, this proposed merger wouldmake the purchase more costly, as Grey-hound’s only choices were Oracle andPeopleSoft. Tr. at 1495:13–21. Withoutthe competition between the two, Gloverforesees prices increasing. Tr. at1495:13–21 (Glover).

Phillip Maxwell, Senior Vice Presidentand Chief Information Officer of the Nei-man Marcus Group (NMG), testified aboutthe ERP needs of the specialty retailer.NMG has properties located throughoutthe country, approximately 15,000 employ-ees and $3 billion in annual sales. Tr. at652:3–13 (Maxwell). NMG formerly hadused FMS software that was originallyfrom MSA, a vendor purchased by Dun &Bradstreet and then GEAC subsequent toNMG’s installation of the software. Tr. at


655:15–22 (Maxwell). In 2002, NMG de-cided to replace its FMS software andbegan conferring with individuals in itsbusiness and technology units, three con-sulting firms and the Gartner Group. SeeTr. at 662:1–663:11 (Maxwell). After ex-amining vendors’ functionality, experiencein retail, price and size/stability, NMG nar-rowed its choices to Oracle and PeopleSoft.Tr. at 665:6–20 (Maxwell). NMG did notconsider SAP because of SAP’s lack ofstrong presence in the retail vertical andMaxwell’s opinion that SAP is ‘‘very ex-pensive to implement.’’ Tr. at 669:11–16(Maxwell). Had the cost of Oracle or Peo-pleSoft FMS software been 10 to 20 per-cent higher, NMG would not have consid-ered SAP, any other FMS vendor, legacysoftware or internally developed software.Tr. at 669:17–670:15 (Maxwell). Based onprice, a high level comparison and detailedGAP analysis, NMG eventually selectedOracle to provide it with FMS software.See Tr. at 671:8–673:2 (Maxwell).

NMG also began licensing HRM soft-ware from Oracle in 2003, though it hasnot yet begun to implement that software.See Tr. at 674:9–11, 676:14–18 (Maxwell).NMG went through a similar process inevaluating HRM software as it did in eval-uating FMS software. Tr. at 684:25–685:20 (Maxwell). As with the FMS soft-ware, NMG concluded that Oracle andPeopleSoft were its only viable alterna-tives. See Tr. at 686:13–16 (Maxwell).NMG did not believe that SAP suited itsneeds as a retailer. See Tr. at 686:11–18(Maxwell). Had the cost of the Oracle orPeopleSoft HRM software been 10 to 20percent higher, NMG would not have con-sidered other HRM vendors, legacy soft-ware, internally developed software or out-sourcing. Tr. at 686:19–687:13 (Maxwell).NMG eventually selected the Oracle HRMsoftware, but based on a 70 to 80 percenthigher target price than previously pre-dicted, NMG has delayed implementationof the Oracle HRM software to look for

cost-reducing options. Tr. at 676:19–677:13 (Maxwell). But Maxwell testifiedthat, even with the 80 percent price in-crease, NMG has not abandoned the Ora-cle HRM. Tr. at 677:20–25 (Maxwell).

Laurette Bradley, Senior Vice Presidentof Information Technology at Verizon, tes-tified about Verizon’s current procurementof new HRM software. Tr. at 577 (Brad-ley). Verizon is a telecommunicationscompany with a ‘‘majority holding in fourof five different countries.’’ Tr. at 580:22–25 (Bradley). Verizon has minor invest-ments in over 30 countries worldwide withan annual revenue of approximately $66billion. Id. Bradley testified that 49 per-cent of Verizon’s labor is unionized world-wide, which places ‘‘significant demandsupon [the] ERP systems, particularly [the]HR and payroll systems’’ because eachunion contract, from each jurisdiction,must be reflected and managed regardingpayroll, vacation, absences, and personaldays. Tr. at 583:6–15 (Bradley). Prior toOctober 2003, Verizon had used two differ-ent HRM programs, one from PeopleSoftand one from SAP. Tr. at 583:23 (Bradley).The PeopleSoft HRM was used to managethe former BellAtlantic part of the compa-ny and SAP HRM was used to manage theformer GTE part of the company. Tr. at584:1–4 (Bradley). The same is true ofVerizon’s FMS. But in October 2003, Veri-zon decided to consolidate the two systemsas far as HRM software. Tr. at 584:11–12(Bradley). Verizon chose PeopleSoftHRM for the entire company and as of thedate of the trial, the new software wasbeing implemented. Id. Bradley testifiedthat a merger between Oracle and People-Soft makes her very concerned that Oraclewill not be interested in upgrading or fur-ther ‘‘developing’’ current PeopleSoft soft-ware. Tr. at 592:5, 593:3–10 (Bradley).Bradley does not want to lose the constant‘‘care, feeding, repair, and evolution’’ thatPeopleSoft now offers to its customers.


Tr. at 592:17–18 (Bradley). When askedwhat other vendors Verizon could turn toin obtaining FMS and HRM that meetVerizon’s complex and international needs,Bradley listed only Oracle, PeopleSoft andSAP. Tr. at 598:7–8 (Bradley). But Brad-ley did testify that Verizon is ‘‘constantly’’considering outsourcing its entire HRmanagement, but so far has determinedthat the risks are just too high. Tr. at604:20–21 (Bradley).

Bradley admitted that Verizon alreadyoutsources its 401(k) stock plans and medi-cal and dental benefits. Tr. at 604:12–14(Bradley). Finally, Bradley stated that ifOracle, PeopleSoft or SAP increasedprices by 10 percent, Verizon would notturn to any other vendors for their FMSand HRM. Tr. at 606:23–25, 607:1–3 (Brad-ley). Further, Verizon would not use itsoff-shore information technology staff todevelop an in-house FMS or HRM systemin response to a 10 percent increase. Tr.at 607:12–15 (Bradley).

Finally, Scott Hatfield, Chief Informa-tion Officer of Cox Communications dis-cussed his company’s ERP software needs.Tr. at 87:8–11 (Hatfield). Cox is the thirdlargest cable television operator in theUnited States, delivering video service toabout six and half million households. Tr.at 89:11–14 (Hatfield). Cox has a presencein 30 states and about 21,000 employees.Cox has annual revenues of over $6 billion.Tr. at 89:22–25 (Hatfield). Hatfield testi-fied that Cox uses PeopleSoft HRM tohandle payroll, recruitment, benefits pro-grams and training. Tr. at 94:14–19 (Hat-field). In 1995, during the HRM vendorprocurement process, Cox only consideredOracle and SAP as other potential vendorsof HRM. Tr. at 96:12 (Hatfield). Hatfieldtestified that while Cox had consideredoutsourcing its HRM altogether, it haddecided against doing so because the com-pany needed to have a ‘‘tight integration’’between its HRM and CMS, which could

not be outsourced. Tr. at 97:17–19 (Hat-field).

Regarding FMS, in 2003, Cox decided tochange from J D Edwards to a new ven-dor. Cox hired Accenture to consult inthis process. Tr. at 114:22–25 (Hatfield).Accenture gave Cox a list of three vendorsof FMS that could meet Cox’s needs: Ora-cle, PeopleSoft and SAP. Tr. at 115:9–10(Hatfield). Hatfield stated that no otherfirms were ‘‘brought to his attention.’’ Tr.at 121:18 (Hatfield). Cox eliminated SAPbecause no one in the company had anyreal experience with SAP and Hatfield didnot want to ‘‘be starting from scratch.’’Tr. at 118:3 (Hatfield). Hatfield statedthat Cox wanted Oracle and PeopleSoft toknow they were the final two competingfor Cox’s FMS business and that Coxasked the two vendors to give their bestprices. Tr. at 126:1–3 (Hatfield). Coxultimately chose Oracle as its FMS vendorbased upon highest level of functionalityratings. Tr. at 129:1–5 (Hatfield). Final-ly, Hatfield stated that if Oracle or People-Soft’s prices had been 10 percent higher,Cox would not have turned to Lawson,Great Plains, best of breed solutions, out-sourcing or writing its own FMS software.Tr. at 136:14–138:23.

In the main, and contrary to the charac-terization of plaintiffs’ counsel before trial,the court found the testimony of the cus-tomer witnesses largely unhelpful to plain-tiffs’ effort to define a narrow market ofhigh function FMS and HRM. Each ofthese witnesses had an impressive back-ground in the field of information technolo-gy. They appeared knowledgeable andwell informed about their employers’ ERPneeds and resources. And the court doesnot doubt the sincerity of these witnesses’beliefs in the testimony that they gave.What the court questions is the groundsupon which these witnesses offered theiropinions on the definition of the product


market and competition within that mar-ket.

[12] The test of market definitionturns on reasonable substitutability. E. I.du Pont, 351 U.S. 377, 76 S.Ct. 994, 100L.Ed. 1264. This requires the court todetermine whether or not products have‘‘reasonable interchangeability’’ basedupon ‘‘price, use and qualities * * *.’’ Id.at 404, 76 S.Ct. 994. What, instead, thesewitnesses testified to was, largely, theirpreferences.

Customer preferences towards oneproduct over another do not negate inter-changeability. See R R Donnelley & SonsCo., 120 FTC 36, 54 n. 65 (1995) (citingRobert Pitofsky, New Definitions of theRelevant Market and the Assault on Anti-trust, 90 Colum. L. Rev. 1805, 1816 (1990)(‘‘There will almost always be classes ofcustomers with strong preferences * * *but to reason from the existence of suchclasses to a conclusion that each is entitledto * * * a separate narrow market defini-tion grossly overstates the market powerof the sellers.’’)). The preferences of thesecustomer witnesses for the functional fea-tures of PeopleSoft or Oracle products wasevident. But the issue is not what solu-tions the customers would like or preferfor their data processing needs; the issueis what they could do in the event of ananticompetitive price increase by a post-merger Oracle. Although these witnessesspeculated on that subject, their specula-tion was not backed up by serious analysisthat they had themselves performed orevidence they presented. There was little,if any, testimony by these witnesses aboutwhat they would or could do or not do toavoid a price increase from a post-mergerOracle. To be sure, each testified, with akind of rote, that they would have nochoice but to accept a ten percent priceincrease by a merged Oracle/PeopleSoft.But none gave testimony about the cost ofalternatives to the hypothetical price in-

crease a post-merger Oracle would charge:e.g., how much outsourcing would actuallycost, or how much it would cost to adaptother vendors’ products to the same func-tionality that the Oracle and PeopleSoftproducts afford.

If backed by credible and convincingtestimony of this kind or testimony pre-sented by economic experts, customer tes-timony of the kind plaintiffs offered canput a human perspective or face on theinjury to competition that plaintiffs allege.But unsubstantiated customer apprehen-sions do not substitute for hard evidence.

While listening to the testimony of thesecustomer witnesses, it became clear to thecourt that these witnesses represent agroup of extremely sophisticated buyersand users of information technology; theyhave decades of experience in negotiatingin this field. This made more evident thefailure of these witnesses to presentcost/benefit analyses of the type that sure-ly they employ and would employ in as-sessing an ERP purchase. The evidenceat trial established that ERP customershave choices outside the integrated suitesof Oracle, PeopleSoft and SAP. Indeed,Glover’s testimony showed that—as Oraclecontends—customers have some leverageby virtue of their existing installed base‘‘to do nothing’’ and thereby resist anti-competitive price increases by ERP ven-dors. Although the court is not convincedthat this is a long-term option due to theever changing business and legal environ-ment in which enterprises operate, thisoption does afford ERP customers somelimited protection and leverage. At anyrate, plaintiffs’ customer witnesses did not,in their testimony, provide the court withdata from actual or probable ERP pur-chases and installations to demonstratethat the witnesses’ employers would havehad no choice but to submit to a SSNIPimposed by a post-merger Oracle.


The court, therefore, finds that thesewitnesses did not establish by a prepon-derance of the evidence that the productsoffered by Oracle, PeopleSoft and SAP arein a distinct line of commerce or productmarket from those offered by other ERPvendors. The court finds that these wit-nesses did not establish that it was morelikely than not that customers of a post-merger Oracle would have no choice but tosubmit to a small but significant non-tran-sitory price increase by the merged entity.These findings do not rest alone on thecourt’s skepticism about the testimony ofplaintiffs’ customer witnesses.

Oracle, too, presented customer wit-nesses, although a much smaller numberof such witnesses. Brian Mearns, Directorof Personnel Service Delivery for Bank OfAmerica (BA), testified about BA andFleet Boston’s (Fleet) needs and decisionsregarding HRM and FMS software. Tr.at 3276:2–21 (Mearns). In April 2004, BAacquired Fleet. Tr. at 3276:10 (Mearns).Mearns had held the title of Director ofHR Service Delivery at Fleet prior to theacquisition. Mearns stated that Fleet hadpersonnel of over 50,000 worldwide, withinvestment and mortgage offices in 32states and throughout South America, Eu-rope and Asia. Tr. at 3280:14–3281:11(Mearns). Mearns testified that Fleet hadused PeopleSoft HRM software since 1996.Tr. at 3286:18–20 (Mearns). In 2002, Fleetsought to upgrade its PeopleSoft HRMsoftware to encompass increased function-ality. But the $12 million price tag wastoo much for Fleet’s appropriation commit-tee and Mearns was told that upgradingPeopleSoft was not an option. Tr. at3289–3290:11. Based upon this turn ofevents, Mearns stated that Fleet insteadturned to outsourcing to meet its HRMneeds. Tr. at 3290:24–25 (Mearns). Thesearch to find an outsourcing firm thatcould meet all of Fleet’s needs led to fivecandidates: Mellon, Hewitt, Exult, Accen-ture and Fidelity. Tr. at 3293:1–2

(Mearns). Fidelity ‘‘best met [the] busi-ness objectives and selection criteria’’ thatFleet required. Tr. at 3295:11–12(Mearns). After implementation of thenew outsourcing solution, Mearns statedthat Fidelity’s systems were ‘‘very config-urable to meet [Fleet’s] requirements.’’Tr. at 3297:12–14 (Mearns). After BA ac-quired Fleet, Mearns gave a presentationto BA executives about Fleet’s experiencewith outsourcing and the capability of Fi-delity. Tr. at 3300:14–17 (Mearns).Based upon this presentation, BA decidedto outsource all of its HRM functions toFidelity. Tr. at 3300:20–22 (Mearns).

Charles Peters, Senior Executive VicePresident for Emerson Electric Company(Emerson), was also called by Oracle totestify about other viable substitutes tohigh function ERP. Emerson is a globalmanufacturing company operating in sixindustries including climate technologies(air conditioning and heating components),motor and appliance components and com-ponents for large industrial equipment.Tr. at 1190–1191:15 (Peters). Emerson’sannual revenue exceeds $15 billion and itsworkforce includes about 110,000 employ-ees in over 50 countries. Tr. at 1191:18–25(Peters). Within these six industries, Em-erson has over 40 divisions. Tr. at 1193:11(Peters). Some of these divisions, stand-ing alone, have global operations and reve-nues in the billions of dollars. Tr. at1193:19–20 (Peters). Many of these divi-sions operate their own HRM and FMSsoftware. Tr. at 1198:7–8 (Peters).

One aspect of Peters’ job is to provide‘‘options’’ to each division regarding theirchoices for handling FMS and HRMneeds. Tr. at 1197:6–18 (Peters). Petersstated these ‘‘options’’ include ERP ven-dors, outsourcing, best of breed solutions,in-house solutions and extending incum-bent systems. Tr. at 1198:7–19 (Peters).Peters testified that one of his divisions


will not implement Oracle ERP becausetheir in-house software fully meets itsneeds. Tr. at 1211:1–18 (Peters). Fur-ther, Peters discussed the increasing rolethat outsourcing to Asia or the Philippinesplays in the HR area of many divisions.Tr. at 1214:7–16 (Peters). Finally, Petersstated that he did not believe that Emer-son divisions would have to pay more forOracle ERP if the proposed merger isconsummated. Tr. at 1235:11–14 (Peters).

In so testifying, Peters cited to a recentnegotiation he conducted with Oracle con-cerning ERP for one division. During thenegotiations, Peters stated, PeopleSoft wasnever a contender. Tr. at 1235:16 (Pe-ters). The possibility of using PeopleSoftwas not leverage that Peters could use toadvantage in seeking to obtain a lowerprice from Oracle. Emerson still receiveda competitive price from Oracle. Tr. at1235:18–24 (Peters). Accordingly, Petersstated that he does not believe that thepresence or absence of PeopleSoft is afactor that constrains Oracle pricing. Id.

To be sure, the testimony of the Oraclewitnesses, like that of the plaintiffs’ cus-tomer witnesses, entailed some speculationabout the presence or absence of People-Soft in the market. But the Oracle wit-nesses testified about concrete and specificactions that they had taken and been ableto complete in order to meet their firms’information processing needs, apart fromrelying on the three ERP vendors thatplaintiffs contend are a market unto them-selves. Hence, the court finds on thisbasis, as well as an assessment of thewitnesses’ credibility, that the testimony ofthe Oracle customer witnesses was morebelievable than that of the plaintiffs’ wit-nesses, despite the greater number of thelatter.

Plaintiffs’ Expert: Iansiti

In addition to the customer witnesses,plaintiffs presented the expert opinion tes-timony of Marco Iansiti, a professor of

business administration at the HarvardBusiness School. Iansiti’s expertise lies inoperations management and informationtechnology. Iansiti also has experience asa consultant for companies in the ‘‘soft-ware space.’’ Tr. at 2020:24 (Iansiti).Iansiti thus brought an academic perspec-tive that basically echoed the testimony ofthe customer witnesses. The court is sat-isfied that Iansiti is well qualified to opineon features of ERP products.

Iansiti was asked to identify the vendorswhose ERP products would meet theneeds of a ‘‘large and complex enterprise.’’Tr. at 2024:4 (Iansiti). Iansiti examinedthe product documents and analysts re-ports of 148 ERP vendors. Tr. at 2025:10(Iansiti). Iansiti testified that only theproducts of Oracle, PeopleSoft and SAPpossess the functionality adequate to meetthe needs of such an enterprise. Withregard to Lawson, Iansiti testified that itsHRM product can handle only three levelsof an organization and its FMS productfive levels and thus is wholly inadequatefor a large and complex enterprise. Tr. at2047:3–5 (Iansiti). By contrast the People-Soft and Oracle products can capture ‘‘un-limited levels of organization.’’ Tr. at2047:17 (Iansiti).

Iansiti testified that Microsoft BusinessSolutions (MBS) provides four ERP prod-ucts: Navison, Axapta, Great Plains andSolomon. Tr at 2054:7–8 (Iansiti). ButMBS sells exclusively through resellersand thus lacks the kind of direct relation-ship necessary to furnish the level andspecific services required by large andcomplex enterprises. Tr. at 2054:17–2055:11 (Iansiti). Microsoft will not, inIansiti’s view, have a single product to‘‘rationalize’’ its present four ERP prod-ucts until 2009. Tr. at 2058:25–2061:10(Iansiti). Iansiti expressed doubts thatMicrosoft will be able to develop productscompetitive with those of PeopleSoft, Ora-


cle and SAP because Microsoft’s businessmodel is radically different from that ofthese three companies. Tr. at 2061:11–2063:15 (Iansiti). Iansiti also saw no de-velopments in internet technology or theintegration layer that would likely replacethe functionalities of the ERP offerings ofPeopleSoft, Oracle and SAP. Tr. at2077:12–2080:11 (Iansiti).

The court finds that Iansiti’s testimonyfails to establish a product market. Iansitidid not claim to have performed an eco-nomic study of the ERP industry. Tr. at2082:5–20 (Iansiti). He conceded thatthere is not a ‘‘clear line or demarcation’’to distinguish enterprises that have highfunctional needs from ‘‘lower function ormid-market needs.’’ Tr. at 2088:7–2090:21(Iansiti). Furthermore, Iansiti concededthat a number of companies that wouldappear to meet the criteria of large andcomplex enterprises have satisfied theirERP requirements with the products ofvendors other than PeopleSoft, Oracle andSAP and have satisfied their needs fromoutsourcing or from their legacy systems.See Tr. at 2091:5–2095:3, 2100:1–2113:15(Iansiti). Because of his lack of economicanalysis and his inability to identify articu-lable product market boundaries (a keyissue in a horizontal merger case), thecourt finds that Iansiti failed to establish aclearly defined product market along thelines alleged by plaintiffs.

Systems Integrators

Plaintiffs presented the testimony of twosystems integrator witnesses in an effortto prove the existence of a separate highfunction ERP market. One of these wit-nesses, Perry Keating of BearingPoint,however, rebutted as much as supportedplaintiffs’ positions regarding market par-ticipants and likelihood of entry into themarket.

Keating is the Senior Vice President ofBearingPoint, one of the largest consultingcompanies in the world. Tr. at 857:12–15

(Keating). BearingPoint is involved in‘‘management consulting’’ which includesthe ‘‘implementation of financial [and] hu-man resource * * * solutions.’’ Tr. at858:4–7 (Keating). At the outset of histestimony, Keating made clear that Bear-ingPoint has taken no position either for oragainst the proposed merger. Tr. at858:11–18 (Keating). Keating stated thatBearingPoint ‘‘wishes both [Oracle andPeopleSoft] well.’’ Id.

Keating started off by supporting plain-tiffs’ product market definition, statingthat BearingPoint’s ‘‘large clients, whetherit be commercial or public service * * *predominant[ly] * * * choose Oracle, Peo-pleSoft and SAP’’ software. Tr. at 867:10–14 (Keating). Keating called these largecustomers ‘‘Tier 1’’ customers, describingtheir needs with regard to multiple curren-cies, languages and legal systems. Keat-ing stated that ‘‘Oracle, PeopleSoft andSAP are the three clear, you know, playersin the marketplace.’’ Tr. at 870:9–10(Keating).

Further, Keating testified that no othervendor could deliver the degree of func-tionality that these three vendors deliver.Tr. at 871:17–20 (Keating). In support ofthese contentions, plaintiffs introduced aquestionnaire that BearingPoint had com-pleted for the European Commission’s in-vestigation of the merger at bar. Ex.P203 at 1. Keating was personally involvedin preparing the responses to this ques-tionnaire. In one question, the EC askedBearingPoint: ‘‘[Is] there a specific mar-ket for supplying EAS * * * to large com-panies, * * * in which only a few vendorsare active?’’ Id. at 11. BearingPoint re-sponded: ‘‘Yes, there is such a market.The vendors are SAP, Oracle and People-Soft * * *.’’ Moreover, BearingPoint’s re-sponses also stated that it believed innova-tion would be slowed in this market as a


result of the proposed merger betweenOracle and PeopleSoft. Id. at 16.

Once the topic turned to the likelihoodof entry into this marketplace by vendorsother than SAP, Oracle or PeopleSoft,Keating’s testimony began to undermineBearingPoint’s response to the EC. Plain-tiffs directed the court’s attention to aportion of the EC questionnaire pertainingto ease of entry. When BearingPoint wasasked to ‘‘indicate at least three companiesthat are potentially able to enter this [EASfor large companies] market,’’ Bearing-Point had listed Microsoft, Siebel andIBM. Id at 14. Moreover, the responsestated that the only barrier to entry bythese three vendors is ‘‘self choice.’’ Id.But when asked at trial by plaintiffs ifKeating was surprised by Microsoft’s ap-proach to acquire SAP, Keating respond-ed: ‘‘No, * * * Microsoft’s not a companythat plays for second.’’ Tr. at 926:22–24(Keating).

On cross-examination, Oracle delveddeeper. When asked if ‘‘there was anyquestion in [his] mind that Microsoft hasthe ability to develop a scalable product,’’Keating replied ‘‘no.’’ Tr. at 940:13–15(Keating). The following testimony pres-ents a good summary of Keating’s contri-bution regarding the potential entry ofMicrosoft into the high function market:

Question (by Oracle Counsel): They’re[Microsoft] coming aren’t they, to thelarge market space?Answer (Keating): Monday they werealmost there [referring to the SAP ac-quisition revelation].Question: Indeed they were.Answer: I had a conference call with mySAP practice [saying], ‘‘you guys mightwant to get new letterheads.’’ I don’tmean to be flip, but it was pretty clearthey’re coming.

Tr. at 942:14–19 (Keating).Furthermore, Keating’s testimony

makes it appear that BearingPoint is roll-

ing out the red carpet for Microsoft’s ar-rival. At trial, an ‘‘alliance’’ between Mi-crosoft and BearingPoint came to lightunder which BearingPoint has agreed tobecome Microsoft’s ‘‘go to partner’’ in thehigh function ERP software market forcustomers that have less than $2 billion inannual revenues. Ex. D5051 at 2.

In the main, the court found Keating’stestimony to be credible. Most particular-ly, Keating’s testimony of the alliance be-tween his company and Microsoft substan-tiates Oracle’s contention that Microsoft isa competitor for much ERP business andable to extend its reach into an arena inwhich plaintiffs contend that only Oracle,PeopleSoft and SAP now compete. Keat-ing’s testimony gives evidence that Micro-soft’s entry into competition may beachieved by a business model differentfrom that followed by Oracle, PeopleSoftor SAP. Microsoft’s ERP products throughthis collaboration with BearingPoint can becustomized, configured and adapted to becompetitive with the offerings of the threecompanies that plaintiffs contend make upthe market, at least up to a level that iswell within the large, complex level of cus-tomer demand that plaintiffs contend re-quires high function ERP.

Nancy Ellen Thomas, the Global andAmericas Financial Management SolutionsLeader for IBM, also called by plaintiffs,testified about IBM’s role as a consultantto ‘‘large, global complex clients’’ procur-ing FMS software. Unlike BearingPoint,IBM has publicly stated its opposition tothe hostile takeover of PeopleSoft by Ora-cle. Ex. D5240R at 13 (stating that a‘‘successful Oracle bid’’ would be a ‘‘nega-tive for IBM * * * [with] possible impacton strong PeopleSoft [and IBM] alliancerevenue’’ and also considering taking a‘‘proactive stance against the [Oracle/Peo-pleSoft] deal’’). Thomas began by echoingmany of the same views that Keating ex-


pressed in regard to the ERP needs oflarge complex customers, including multi-ple geographies, currencies, languages andregulatory requirements. Tr. at 474:9–12(Thomas). When asked, based upon herexperience, which ERP vendors could offera product that could satisfy the require-ments of these customers across multiplecountries, Thomas listed only Oracle, Peo-pleSoft and SAP. Tr. at 475:2 (Thomas).When asked what vendors could supportreporting requirements for multipleranges of legal entities, Thomas listed onlyOracle, PeopleSoft and SAP. Tr. at 476:3(Thomas). The same three vendors werelisted when Thomas was asked about sup-porting multiple lines of business. Tr. at476:15 (Thomas). Thomas downplayed therole that Lawson plays within this ‘‘upmarket’’ sector, stating that ‘‘the clients* * * we work with are typically not’’ fo-cusing on Lawson to the extent that theyare focusing on Oracle, PeopleSoft andSAP. Tr. at 495:10–15 (Thomas).

Plaintiffs also appeared to use Thomasto bolster their contention on ‘‘localization’’between Oracle and PeopleSoft by askingThomas about the banking industry andwhich firms compete for that business.When asked which vendors she would ex-pect to see in the final scoring and recom-mendation phase of a banking customer’sselection process, Thomas stated: ‘‘primar-ily Oracle and PeopleSoft.’’ Tr. at 498:21–25 (Thomas).

When Oracle’s counsel questionedThomas about the possible bias of IBM,Tr. at 499–503 (Thomas), Thomas admittedthat IBM has the ‘‘largest PeopleSoft prac-tice of any consulting firm in the world’’and that PeopleSoft has ‘‘publicly de-scribed IBM as PeopleSoft’s strongestpartner.’’ Tr. at 499 (Thomas). Further,IBM has over 150 employees dedicated toconsulting and implementing PeopleSoftproducts, all of whom could lose their jobsif PeopleSoft was merged with Oracle, a

company for which IBM has only 75 dedi-cated consultants. Tr. at 500:20–502:10(Thomas).

Turning to Lawson, when asked aboutIBM’s large and complex implementationof Lawson HRM for the State of Arizona,which has over 60,000 employees, Thomasstated that she didn’t have the ‘‘Lawsonexpertise’’ to talk about that transaction.Tr. at 519:12–13. Further, Thomas ‘‘wasnot aware’’ of IBM’s implementation ofLawson software at Montgomery CountySchools in Maryland, an entity with over140,000 students. Nor was she ‘‘aware’’ ofIBM’s implementation of Lawson for theState of Michigan or IBM’s implementa-tion of Lawson for a large school district inTampa. Tr. at 520:7–19 (Thomas).

The court first notes a possible IBMbias due to IBM’s potential loss of People-Soft implementation business, a significantsource of IBM revenue. Furthermore, thecourt cannot overlook Thomas’ lack ofknowledge about any potential high func-tion implementation of Lawson software.This makes the court reluctant to affordmuch, if any, weight to her testimony.Thomas seemed not to be able to identifyfactors that would keep Lawson from com-peting in the high function sector. Hertestimony failed to substantiate plaintiffs’claim of separate FMS and HRM highfunction markets.

Industry Witnesses: PeopleSoftand Microsoft

Next, plaintiffs presented the testimonyof several industry witnesses in an effortto support the proposed high functionERP market.

Richard Bergquist, Chief TechnologyOfficer, Senior Vice President and People-Soft ‘‘Fellow,’’ explained to the court howPeopleSoft defines a high function custom-er versus a mid-market customer. Tr. at255:18–19, 275–276:21 (Bergquist). Notsurprisingly, Bergquist’s definition of high


function customers and high function soft-ware echoed plaintiffs’ definitions (or, atleast, some of them). First, Bergquist stat-ed that a customer cannot be labeled ashigh function simply based upon its size orrevenue. Rather, one ‘‘ha[s] to look all thedifferent dimensions’’ in order properly todistinguish between these two types ofcustomers. Tr. at 276:3 (Bergquist). The‘‘different dimensions’’ that Bergquist re-ferred to in guiding an explorer throughthe task of deciding what label to apply toa customer are: functionality, flexibility,scalability, reliability and technology. Tr.at 280–282, 283:18, 289:4–25 (Bergquist).Only after knowing the customer’s needsregarding all of these dimensions, whichone must learn ‘‘through a series of con-versations with the customer,’’ can onethen place a customer in the correct talis-manic column of high function or mid-market. Tr. at 276:11–13 (Bergquist).

A high function customer requires soft-ware that is highly functional, highly flexi-ble, contains large scalability and is reli-able 24 hours per day, seven days a week.Tr. at 283–289 (Bergquist). Customerswho do not need software with such deepfunctionality, large scalability or high flexi-bility are mid-market customers who buymid-market software. Tr. at 300:10–13(Bergquist). Bergquist succinctly statedthat ‘‘customers that don’t have the needsof large and complex enterprises, we [Peo-pleSoft] group into the mid-market.’’ Tr.at 275:1–2 (Bergquist). Bergquist clearlystated that a market exists for the sale ofhigh function software to high functioncustomers, and in this market PeopleSoftcompetes with only SAP and Oracle. Tr.at 279:17 (Bergquist). Berquist went on toexplain that a customer can be high func-tion regardless of its international loca-tions or international currency needs. Tr.at 292:20 (Bergquist). ‘‘Internationality’’was not a dimension for delineating highfunction from mid-market, rather interna-tional needs simply create the need for

more function and scalability. Nonethe-less, multiple currency, language and na-tionality capabilities are not requirementsfor a high function customer, as a custom-er can be located in the United States onlyand use only English and still be a highfunction customer according to Bergquist.Tr. at 292:1–15 (Bergquist).

Questions soon turned to Lawson and itsrole in this high function software market.Berquist stated that PeopleSoft ‘‘does notbelieve’’ that Lawson sells any HRM orFMS software that has similar functionali-ty to the same software sold by People-Soft. Tr. at 299:21–25 (Bergquist). Rath-er, Lawson has FMS and HRM that is‘‘adequate for the basics of what an organi-zation would need.’’ Tr. at 300:4–5(Bergquist). If the organization has sim-ple and repetitive tasks, then ‘‘the Lawsonproduct does that very well.’’ Tr. at300:9–10, 304:1–4 (Bergquist). But if acustomer starts going beyond those basictasks, then the customer needs featuresand functions that Lawson cannot supply.Tr. at 300:10–13 (Bergquist). Moreover,Lawson does not have to ability to supportUnicode, a common character set for alllanguages of the world. Since Lawsoncannot do that, it is ‘‘limite[d] to the US,Canada and UK.’’ Tr. at 301:19–25(Bergquist).

The next topic was AMS and its role inthe high function market. Bergquist stat-ed that AMS only has a ‘‘financial productthat is meant for sale in the public sector.’’Tr. at 309:14–15 (Bergquist). Further, thesoftware was developed only for a mini-mum level of functionality and requiresextensive customization before it can beimplemented. Tr. at 309:12–17(Bergquist). Further, AMS does not havean HRM product. Because AMS does notrise to the level of functionality required tobe considered high function, AMS is not ahigh function vendor selling a product that


competes in the proposed market. Tr. at310:4–7 (Bergquist).

Next, Bergquist took aim at the best ofbreed solutions, stating that a customer‘‘can’t assemble point solutions to get thefull picture.’’ Tr. at 311:12–14 (Bergquist).These point solutions do not provide corefunctionality, requiring a customer to pur-chase core functionality from a differentvendor, and then having ‘‘multiple solu-tions from point solutions,’’ creating exten-sive integration costs. Tr. at 311:12–25(Bergquist). Accordingly, best of breedsolutions are not a viable option for highfunction customers and therefore are notsubstitutes for high function software.

Next, Bergquist set out to prove thatoutsourcing is also not a viable option forhigh function customers stating that ‘‘wesee it [outsourcing] as less capable soft-ware than that provided by PeopleSoft,SAP and Oracle.’’ Tr. at 314:11–12(Bergquist).

Finally, Bergquist was questioned aboutpotential localized competition between Or-acle and PeopleSoft, thus establishing thelikelihood of unilateral anticompetitive ef-fects. Bergquist testified that there aresome instances where Oracle is People-Soft’s closest competitor over SAP. Tr. at319:6–8 (Bergquist). This type of situationarises in the service industries accordingto Bergquist because Oracle and People-Soft both ‘‘grew up in the same neighbor-hood,’’ the services industry neighborhood,thus making Oracle and PeopleSoft strongcompetitors in this vertical, especiallyamong those who have a ‘‘buy-Americantendency.’’ Tr. at 319:11–16 (Bergquist).Moreover, Bergquist testified that SAPhas suffered from the ‘‘stereotype of Ger-man engineering’’ that leads many to viewSAP software as less flexible and requiringmore customization. Tr. at 320:11–15(Bergquist). But Oracle and PeopleSoftare both seen as very flexible, again mak-

ing them more likely competitors overSAP. Tr. at 320:16–18 (Bergquist).

On cross, Bergquist was first askedabout the distinction between mid-marketand high function customers and software.When asked if there were any PeopleSoftdocuments which describe this distinctionbetween high function and mid-marketcustomers or software, Bergquist said thathe was not aware of any such documents.Tr. at 347:22–25 (Bergquist). Further,Bergquist admitted that there are no‘‘clear-cut answers’’ or ‘‘firm dividing lines’’that distinguish a mid-market customerfrom a high function customer. Tr. at353:15–22 (Bergquist).

Next Bergquist was asked about his dis-missal of Lawson from the high functionmarket. When asked if PeopleSoft hadlost any business from large and complexcustomers to Lawson, Bergquist replied:‘‘I can’t think of any that we have * * *lost.’’ Tr. at 364:5 (Bergquist). Oraclethen showed Bergquist a document, creat-ed by PeopleSoft, tabulating enterprisedeals which PeopleSoft had competed for,and the name of the competitor on thedeal. Ex. D6236A. The data read thatLawson was an enterprise competitor 27times, with SAP competing 33 times andOracle 38 times. Id at PS–C077332. ButBergquist stated: ‘‘I don’t know anythingabout this document * * * where it camefrom or how it was.’’ Tr. at 375:3–9(Bergquist).

Bergquist was then asked about specificinstances of competition with Lawson.When asked if he knew anything aboutPeopleSoft’s loss to Lawson on the DeanFoods account, Bergquist stated ‘‘no.’’ Tr.at 377:16–18 (Bergquist). When askedabout PeopleSoft’s loss to Lawson on theQwest [transcript misspelling] Communi-cations account, again Bergquist statedthat he knew nothing about that lost busi-ness. Tr. at 377:19–21 (Bergquist).


Moreover, it appears Bergquist was noteven aware of instances in which People-Soft won business when in competitionwith Lawson. When asked if he knewanything about PeopleSoft’s wins overLawson on the Maricopa County account,Bergquist replied ‘‘no.’’ Tr. at 377:22–24(Bergquist). Bergquist provided the sameanswer when asked about PeopleSoft’s winagainst Lawson on the San Diego UnifiedSchool District account. Tr. at 378:2(Bergquist). Bergquist, like Ms Thomasbefore him, seemed to have been struckwith a singular memory lapse. It appearsboth witness, while able to testify thor-oughly about other vendors, drew a com-plete blank when asked about potentialhigh function implementations of Lawson.The court began to wonder if this phenom-enon, perhaps called ‘‘Lawson Amnesia,’’would strike any more of plaintiffs’ wit-nesses.

The final part of Bergquist’s cross camewhen defense counsel began inquiringabout the alleged localization of competi-tion between PeopleSoft and Oracle in theservices industry vertical:

Question (Oracle counsel): Can youidentify for me any particular verticalsin which you believe that SAP is notcompetitive with Oracle and PeopleSoft?Answer (Bergquist): SAP may competein almost all the verticals that are there.* * *. There is relative strength forPeopleSoft and Oracle in the servicesindustries.Question: I understand that you’ve saidthat sir, but my question is different.In any of those services industries, is ityour testimony that SAP is not compet-itive with Oracle and PeopleSoft?Answer: No.

Tr. at 388:1–11 (Bergquist) (emphasis add-ed).

Notwithstanding any bias, Bergquist’stestimony served to hurt plaintiffs’ claimsrather than bolster them. First,

Bergquist conceded that no ‘‘clear-cut’’ di-viding line exists in labeling a customer as‘‘high function’’ rather than ‘‘mid-market.’’Finding an articulable division between so-called high function and mid-market ERPis necessary to plaintiffs’ burden of estab-lishing a product market. Second,Bergquist conceded that there is not onesingle services industry vertical in whichSAP is not ‘‘competitive’’ with Oracle andPeopleSoft. The court must demarcatesuch a ‘‘node’’ or area of localized competi-tion between Oracle and PeopleSoft as aprerequisite to finding any likelihood ofunilateral anticompetitive effects.Bergquist’s testimony was also full of self-serving statements regarding the low func-tionality of AMS and Lawson, testimonythat was shown to be wholly unreliable oncross-examination when Bergquist wasrendered unable to remember key infor-mation regarding Lawson.

Philip Wilmington, Executive Vice Pres-ident of PeopleSoft Americas, further tes-tified in support of the plaintiffs’ proposedproduct market. Tr. at 1760:4 (Wilming-ton). Wilmington began by expoundinghow PeopleSoft characterizes the mid-market versus the ‘‘up-market’’ or highfunction market. Tr. at 1765–1766 (Wil-mington). Wilmington stated that the‘‘up-market’’ is defined as customers thathave revenues of $1 billion or above andhave ‘‘complex requirements.’’ Tr. at1765:16–22 (Wilmington). Prior to thePeopleSoft acquisition of J D Edwards,the demarcation line between mid-marketand up-market had been $500 million. Tr.at 1847:7–17 (Wilmington). Predictably,Wilmington stated that PeopleSoft onlycompetes in the up-market with Oracleand SAP. Tr. at 1773:14 (Wilmington).Oracle and SAP are the ‘‘ones [PeopleSoft]runs into all the time.’’ Tr. at 1773:19–20(Wilmington). Only these vendors have‘‘the functionality’’ and the ‘‘references orcustomer successes’’ which allow them to


be a competitive presence in the up-mar-ket. Tr. at 1773:21–1774:2 (Wilmington).Wilmington further testified that whenthese three competitors compete, it gets‘‘very aggressive.’’ Tr. at 1797:20 (Wil-mington). Moreover, Wilmington statedthat oftentimes PeopleSoft knew the iden-tity of its competitors on any given ac-count, with that information driving higherdiscounts. Wilmington cited the exampleof the Oracle and PeopleSoft competitionfor Target, in which Target would commu-nicate the other competitor’s discount of-ferings to PeopleSoft. Tr. at 1797:20–25(Wilmington).

Wilmington testified that he did not be-lieve that the ‘‘do nothing’’ option was athreat to PeopleSoft or other up-marketvendors. Wilmington stated: ‘‘Almostnever do I see a company that is investedin a [procurement] evaluation * * * justdo nothing.’’ Tr. at 1792:2–3 (Wilmington).Wilmington stated that incumbent systemsmay simply ‘‘delay the decision’’ to buyERP, but it is not a long-term solution forany customer. Tr. at 1792:21–22 (Wil-mington).

Testimony turned to Lawson and itsclassification as a mid-market or up-mar-ket vendor. ‘‘Very, very infrequently do Isee Lawson,’’ stated Wilmington in de-scribing the competition for high functioncustomers. Tr. at 1803:9 (Wilmington).‘‘They are not a viable competitor for theup-market.’’ 1803:10–12 (Wilmington).Wilmington stated that Lawson competes,and competes well, in the mid-market sec-tor, and perhaps it can be seen sporadical-ly in the up-market healthcare and retailindustry. Tr. at 1805:13–23 (Wilmington).When Wilmington was asked about thecompetition between PeopleSoft and Law-son on the Amerigroup account, Wilming-ton stated that Americgroup ‘‘was verymuch a mid-market opportunity.’’ Tr. at1810:5–6 (Wilmington). Regarding Micro-soft, Wilmington stated that PeopleSoft

does not compete with Microsoft in the up-market and only sees it from ‘‘time totime’’ in the mid-market. Tr. at 1811:14(Wilmington).

Next, Wilmington was asked about out-sourcing and its role in the up-market.Tr. at 1812:13–14 (Wilmington). Wilming-ton stated that he does not see outsourcingas a threat to PeopleSoft; rather, he seesoutsourcing as an opportunity. Tr. at1812–17–18 (Wilmington). Wilmingtonstated that outsourcers have to buy soft-ware to manage the client’s HR needs, andPeopleSoft tries to be the vendor to supplysuch software. Tr. at 1812:17–18 (Wil-mington). Accordingly, outsourcing is abusiness opportunity, not a threat. Whenasked about outsourcers who use their ownsoftware to manage HR, Wilmington stat-ed that he doesn’t feel threatened becausethat software lacks the ‘‘robust functionali-ty that is going to be necessary to success-fully meet the needs of [the] up-market.’’Tr. at 1813:12–14 (Wilmington).

Finally, Wilmington testified regardinglocalized competition between Oracle andPeopleSoft. Wilmington stated that SAPsoftware was ‘‘developed for a more rigidbusiness model’’ and therefore lacks flexi-bility. Tr. at 1815:5–6 (Wilmington). Ora-cle and PeopleSoft possess such flexibilityand therefore are better solutions for up-market customers. Tr. at 1815:11–15 (Wil-mington). Moreover, Wilmington testifiedthat he believes SAP is more expensive,ranging anywhere from ‘‘20 to 50 percent,in terms of higher cost of ownership acrossthe board.’’ Tr. at 1817:5–7 (Wilmington).In fact, Wilmington cited one example, thePNC Bank account, in which Oracle, SAPand PeopleSoft were all three competing.But SAP was eliminated because its soft-ware did not possess the flexibility thatPNC required. Accordingly, PNC nar-rowed the competition to only Oracle andPeopleSoft. Finally, Wilmington stated


that the PNC scenario was exemplary ofthe situation in the entire United Statesbanking industry. Tr. at 1817–1818 (Wil-mington).

On cross, Wilmington was questionedabout PeopleSoft’s up-market versus mid-market demarcation. It was during thisquestioning that the court learned that theday prior to Oracle’s tender offer for Peo-pleSoft, the demarcation line was $500 mil-lion in revenues and/or 2,000 employeesbut, soon thereafter, the of demarcationline increased to $1 billion in revenue only.Tr. at 1848:10–16 (Wilmington). Wilming-ton stated that it was the J D Edwardsacquisition, and not the tender offer, whichcaused the increase. But Oracle’s counselthen asked: ‘‘If you drew the line at 500million and/or 2,000 employees for [the]mid-market [roof amount], then the up-market would include players other thanOracle, PeopleSoft and SAP, isn’t thatright?’’ Tr. at 1849:5–7 (Wilmington).Wilmington had trouble giving a directanswer to this question, choosing insteadto argue that other factors were necessary,other than revenue, before being able toclassify a customer in the mid-market orup-market. Tr. at 1849:8–15 (Wilmington).But Oracle’s counsel quickly brought Wil-mington’s attention to a 2003 PeopleSoftPricing Policy Document which stated that‘‘[a] mid-market customer is defined as acustomer with revenues of up to $500 mil-lion and/or 2,000 employees.’’ Ex. P4965at 6–7. There is no mention of otherfactors such as scalability or functionalityneeds. Further, the document stated that‘‘[t]he revenue-based metric is meant to bethe single determinant of the Mid–MarketLine * * *.’’ Ex. P4965 at 6–7 (emphasisadded).

Regarding Lawson, Wilmington stood byhis deposition statement that not once in25 years had Wilmington seen PeopleSoftcompete with Lawson for a ‘‘enterprisecustomer.’’ Tr. at 1856:21–25 (Wilming-

ton). Wilmington stated that he basedthis statement upon the new $1 billiondemarcation line between mid-market andup-market (enterprise) customers. Tr. at1857:5 (Wilmington). But Wilmington con-ceded that if the $500 million/2,000 employ-ee line were used, then PeopleSoft hadcompeted with Lawson for enterprise cus-tomers. Tr. at 1858:7–8 (Wilmington).Oracle then introduced a document createdby PeopleSoft in July 2003, after the Ora-cle offer and the J D Edwards acquisition,which showed the number of times People-Soft had competed with certain vendors onenterprise deals. Tr. at 1858:10–17 (Wil-mington); Ex. D6236. Since the documentwas created after the J D Edwards acqui-sition, it would appear that the mid-marketdemarcation line used would be (or shouldhave been) the $1 billion line. The docu-ment lists PeopleSoft as having competedwith Lawson 27 times for an enterprisecustomer. Tr. at 1859:7–8 (Wilmington).When asked if Wilmington still stood byhis testimony, Wilmington stated that hestill believed that PeopleSoft ‘‘doesn’t seeLawson in enterprise deals.’’ Tr. at1859:13–14, 1861:5–7 (Wilmington). Wil-mington elected to ‘‘stand by his testimo-ny.’’ Tr. at 1861:8 (Wilmington). LawsonAmnesia appeared to have claimed yet athird victim.

When asked about PeopleSoft’s competi-tion with Lawson for HCA Columbia, Wil-mington could not speak to that issue be-cause ‘‘he had not been involved in thecompetition.’’ Tr. at 1868:4 (Wilmington).When shown a PeopleSoft document thatlisted Lawson as ‘‘the number one compet-itor in new market deals’’ in the westerngeographic region of the United States,Wilmington stated that he ‘‘did not know’’if Lawson was really number one. Tr. at1866:13–21 (Wilmington). When compet-ing with Lawson for the Stanford Univer-sity Medical Center, PeopleSoft documentswritten by Lynn Duffy, the sales team


leader on the deal, stated that ‘‘Lawson isthe competition,’’ but Wilmington statedthat he ‘‘was not sure’’ if Duffy was rightabout that point. Tr. at 1870:5–6 (Wil-mington). When asked about PeopleSoft’sloss to Lawson for the State of Michiganaccount, Wilmington stated that he was‘‘not certain’’ if PeopleSoft had even com-peted against Lawson for that account.Tr. at 1878:15–16 (Wilmington). Whenasked about PeopleSoft’s loss to Lawsonfor ManuLife’s business, Wilmington stat-ed that he ‘‘did not remember losing tothem’’ on that account. Tr. at 1896:14(Wilmington). When asked about the lossto Lawson for the Mayo Clinic account,Wilmington stated that he ‘‘was not famil-iar with the details of that competition.’’Tr. at 1896:22–23 (Wilmington). SinceWilmington apparently was not aware ofwhat PeopleSoft’s own documents revealabout Lawson as a competitor and is ‘‘notcertain’’ whether PeopleSoft competedagainst Lawson for several large accounts,the court finds Wilmington’s testimonyconcerning Lawson’s absence from the up-market largely incredible.

Regarding outsourcing, Wilmington wasshown the same document created by Peo-pleSoft soon after the Oracle tender offer,which showed that PeopleSoft competedagainst ADP, an outsourcer, 15 times. Ex.D6236. Wilmington stated that he ‘‘didnot know how this [sic] data was compiledand edited;’’ therefore, he could not statewhether these data meant that PeopleSoftfaced competition from ADP for up-marketcustomers. Tr. at 1860:5–15 (Wilmington).

Finally, Oracle cross-examined Wilming-ton about any alleged localization betweenOracle and PeopleSoft. Wilmington wasshown his video deposition in which he wasasked:

Question (Oracle counsel): Is there anyvertical segment of the market in theUnited States, the ERP market, whereyou do not consider SAP to be a formi-

dable competitor for large enterprisecustomers?Answer (Wilmington): For large enter-prise customers, no. I believe them tobe a formidable competitor across theindustry.

Tr. at 1957:10–21 (Wilmington) (emphasisadded). When asked if he had given thoseanswers, Wilmington replied ‘‘yes.’’ Tr. at1957:20 (Wilmington).

For the same reasons the court men-tioned above in discounting Berquist’s tes-timony, the court cannot accord muchweight to Wilimgton’s testimony. First,Wilmington admitted that there is not asingle vertical industry in which Wilming-ton does not believe SAP to be a ‘‘formida-ble competitor’’ undercutting plaintiffs’unilateral effects claim. Further, in de-scribing the way in which PeopleSoft char-acterizes mid-market customers, Wilming-ton was impeached by a document createdby his own company. Likewise, the samedocument impeached his testimony aboutthe absence of outsourcers from the up-market. This impeachment, combinedwith Lawson Amnesia, leads the court tofind that Wilmington did not offer reliableevidence establishing an articulable prod-uct market containing only Oracle, People-Soft and SAP.

Douglas Burgum, Senior Vice Presidentof Microsoft Business Solutions (MBS),was another industry witness called byplaintiffs in order to support their theoryof the high function product market and itsthree participants. Burgum began by de-scribing how he literally ‘‘bet the farm’’ ona small software company called GreatPlains in 1983. In 2001 Microsoft acquiredGreat Plains. Tr. at 2974:3–8 (Burgum).In 2002, Microsoft acquired Navison Soft-ware, a Danish company. The entiregroup was rebranded as ‘‘Microsoft Busi-ness Solutions’’ (MBS). Tr. at 2973:8–9(Burgum). MBS sells four business appli-


cation products: Navison, Great Plains,Solomon and Axapta. Tr. at 2996:15–16(Burgum). Burgum is responsible for theoverall performance and market strategiesof MBS, as well as ongoing developmentsof new products and enhancements to ex-isting products. Tr. at 2974:18–20 (Bur-gum). Burgum began by stating thatMBS is focused on selling its product tomid-market customers. Tr. at 2978:5–8(Burgum).

To Microsoft, mid-market customers arecustomers who have employees rangingfrom 50 to 1000 employees and an averageIT spend between $10,000 and $2 million.Ex. P2533R at 6. Further showing thatMicrosoft is only focused on mid-marketcustomers, Burgum testified that MBSdoes not have a sales force. Rather, MBSis sold indirectly through reselling part-ners, companies whose sole purpose is toresell MBS products. Tr. at 2986–2988(Burgum). Moreover, neither partnersnor MBS itself offers implementation orconsulting services for the products and donot intend to do so in the future. Tr. at2995:3–18 (Burgum). When asked if MBSintended to expand its products’ ability toserve the large enterprise sector, Burgumresponded ‘‘no, * * * that is not a segmentwe are targeting.’’ Tr. at 3001:20–3302:1(Burgum). Moreover, MBS products donot have the functionality to meet largecustomers’ needs. Tr. at 3005:22–25 (Bur-gum). MBS products, Burgum stated,cannot handle the ‘‘multi-multi-multi is-sues,’’ such as multiple languages and cur-rencies that large organizations tend toneed. Tr. at 3011:23–25 (Burgum). Whenasked what firms’ software could meetthose needs, Burgum responded: Oracle,PeopleSoft and SAP. Tr. at 3006:8–9 (Bur-gum). Burgum stated that while Micro-soft competes with these three from timeto time, that competition only occurs formid-market customers. Tr. at 3008:3–6(Burgum).

Burgum cited the lost North Dakotaaccount as an example of the limited func-tionality of the Great Plains product, bothpre- and post-acquisition with Microsoft,and its inability to meet large functionalneeds. Tr. at 3022:3–7 (Burgum). Bur-gum was asked why Microsoft didn’t ‘‘justspend a bunch of money’’ to redevelop thecode and the salesforce in order to com-pete for larger accounts. Tr. at 3024:3–10(Burgum). Burgum stated that undertak-ing would ‘‘be a formidable task’’ andwould ‘‘take more money than I would bewilling to recommend that Microsoftspend.’’ Tr. at 3024:12–18 (Burgum).Plaintiffs asked Burgum about the Micro-soft/SAP acquisition proposal. Burgumstated the ‘‘leading’’ reason that Microsoftwanted to acquire SAP was not to enterthe high function market for ERP andthereby start competing with Oracle orPeopleSoft. Rather the acquisition was tocreate ‘‘a better value for the customerswho would use Microsoft Office to workwith and make decisions around the datathat would come out of the SAP system.’’Tr. at 3040:9–15 (Burgum). Microsoft sim-ply wanted to purchase SAP in order tohelp ‘‘front-end users’’ be ‘‘better able tocommunicate with back-end data.’’ Tr. at3039:25–3040:2 (Burgum). See Ex. P841Rat 1. Apparently, the acquisition was notmotivated by any ill-will towards Oracle orany desire to enter the market and beginundercutting Oracle. The discussions be-tween Microsoft and SAP were concludedin early spring 2004, about the time thissuit went to trial, with a decision not tomove forward with the acquisition. Tr. at3028:9–10 (Burgum).

When asked about Microsoft’s alliancewith BearingPoint, Burgum testified thatMicrosoft had only the humblest of inten-tions in entering into this alliance. Underthis agreement, Microsoft was to ‘‘providefunding for hiring, recruiting and trainingof people who would get skilled up on


Axapta.’’ Tr. at 3055:15–17 (Burgum). Inreturn, BearingPoint agreed to recom-mend, install and maintain MBS software,specifically Axapta, to BearingPoint con-sulting clients. Tr. at 3055:15–17 (Bur-gum); Ex. 3249R at 4, 15. MBS had noplan or expectation for BearingPoint torecommend Axapta software to high func-tion customers. Tr. at 3053:1–6 (Burgum).To the contrary, this agreement was onlyentered into for BearingPoint to sell MBSto mid-market customers. Tr. at 3054:10–18 (Burgum).

Subsequent to trial, BearingPoint an-nounced that the new Microsoft BusinessSolutions Axapta was ‘‘a compelling ERPsolution’’ which ‘‘provides functionalityacross all key areas of the business * * *including financial management, CRM[and] HR management * * *.’’ SeeBearingPoint homepage at http://www.bearingpoint.com/solutions/enter-prise solutions/microsoft bus sol.html.The BearingPoint webpage claims the‘‘key’’ functionalities of Microsoft Axaptainclude ‘‘multiple companies, multiple lan-guages, and multiple currencies.’’ Id. Al-though these statements do not appear inthe trial record, they are consistent withthe substantial evidence in the record andafford additional reason to discount Bur-gum’s testimony that MBS is not at leasta potential entrant in what plaintiffs char-acterize as the high function market.

The court accords little weight to Bur-gum’s testimony attempting to prove Mi-crosoft’s absence from the so-called highfunction ERP product market. Burgum’sUriah Heep like humility about Microsoft’sintentions regarding the failed SAP alli-ance and the successful BearingPoint alli-ance was unconvincing. It strains creduli-ty to believe that Microsoft would offerbillions of dollars to acquire SAP merely tomake data processing easier for customerswho use both Microsoft Office and SAP

ERP. Further, this proposition is im-peached by Microsoft’s actions with Bear-ingPoint concurrently, or soon after, theSAP alliance was discontinued. Finally,the court wholly discounts Burgum’s testi-mony that MBS software, especially Axap-ta, lacks the functionality to be consideredhigh function ERP. Burgum stated thatMBS products cannot provide the ‘‘multi,multi, multi’’ functionality, but Bearing-Point is selling Axapta on the basis that itcan handle ‘‘multiple languages, currenciesand businesses.’’ Accordingly, the courtdiscounts Burgum’s testimony portrayingMBS solely as a mere humble mid-marketvendor.

Finally, in attempting to show the highbarriers to entry into the high functionmarket, plaintiffs called Richard Allen, for-mer Executive Vice President of FinanceAdministration and CFO of J. D. Edwards.Tr. at 747:20–25 (Allen). Allen testifiedthat J. D. Edwards, prior to being ac-quired by PeopleSoft, had been a company‘‘focused on mid-market customers’’ thatdid not need ‘‘high levels of configurabili-ty,’’ ‘‘deep functionality’’ or high scalability.Tr. at 746:20–21, 757:25–758:20 (Allen).But Allen testified that in the early 1990sJ. D. Edwards attempted to reposition it-self in order to sell to ‘‘up-market custom-ers.’’ Tr. at 770:22–771:10 (Allen). J. D.Edwards had to ‘‘create a software archi-tecture to allow [its] products to run onmultiple hardware platforms, with multipledatabases and multiple operating sys-tems.’’ Tr. at 771:16–21 (Allen). But J. D.Edwards ultimately abandoned this at-tempted repositioning in 2001. Tr. at777:6 (Allen). Allen stated that ‘‘[J. D.Edwards] came to the conclusion that afterabout a decade involved in the effort, hun-dred of millions of dollars of investment,we didn’t have the products, services, andultimately the reputation necessary to sat-isfy the requirements that up-market cus-tomers have.’’ Tr. at 777:8–13 (Allen).


Accordingly, plaintiffs argue that no firmcould enter the high function FMS andHRM markets within the required twoyear threshold set by the Guidelines. Pls.Post Brief (Doc. # 366) at 44 (citing Guide-lines § 3.3).

Plaintiffs’ Expert: Elzinga

By far the most important of plaintiffs’witnesses was Professor Kenneth Elzingaof the University of Virginia. Elzinga is awell known and highly regarded econo-mist. Tr. 2142–2145 (Elzinga); Ex. 4014A.The court finds Elzinga to be highly quali-fied to offer testimony on market defini-tion. Elzinga was the only one of plain-tiffs’ witnesses who offered testimony fromwhich the court could attempt to calculatemarket shares and apply the PhiladelphiaNat Bank presumptions or perform theHHI calculations of the Guidelines.

In reaching his proposed market defini-tion, Elzinga purported to follow theGuidelines approach. Tr. at 2163:18–19(Elzinga). Elzinga concluded that the rel-evant market was limited to high functionFMS and HRM software. Elzinga testi-fied that a hypothetical monopolist couldprofitably impose a SSNIP in high func-tion FMS and HRM. Tr. at 2149:16–22(Elzinga). Elzinga posited that if amerged Oracle/PeopleSoft decided to in-crease the price of its high function FMSand HRM products, consumers would notsubstitute (1) mid-market solutions (suchas those produced by ERP vendor Law-son), (2) best-of-breed solutions (such asthose produced by vendor Kronos), (3) in-cumbent or legacy solutions or (4) theservices of outsourcing firms (such as Fi-delity and ADP). Tr. at 2178:10, 2179:8–14(Elzinga).

Elzinga reached his conclusions by ana-lyzing several ‘‘strains’’ of evidence: (1)Oracle discount approval forms; (2) re-ports from independent research firms;(3) information from high function FMSand HRM customers and consulting firms;

and (4) internal documents from firms inthe enterprise software sector. Tr. at2168:9–11, 2180:4, 2184:10–12, 2188:23–25(Elzinga). This being established, Elzingathen presented his conclusions on marketshares. Tr. at 2209–2220 (Elzinga).

Discount approval forms. Elzinga’sfirst strain of evidence, and the one onwhich he appeared to place the greatestemphasis, was the analysis and tabulationof Oracle discount approval forms (DAF).See Exs. P1000–P1944.

Oracle salespersons have the discretionto offer a 20 to 25 percent discount off thelist price of HRM or FMS. Tr. at 2168:25–2169:2 (Elzinga). If a situation arises inwhich Oracle is competing with anotherERP vendor and this requires the Oraclesalesperson to offer a larger discount onthe Oracle ERP software, a DAF must beexecuted and approved by an Oracle offi-cial. Tr. at 2169:3–9 (Elzinga). In exe-cuting a DAF, the salesperson lists the‘‘justification’’ for pursuing an increaseddiscount. In the justification column ofthe DAF, ‘‘sometimes the particular com-petitor or alternative [solution] that is jus-tifying * * * or provoking the discountthat is being requested’’ can be found.Tr. at 2169:17–19 (Elzinga).

Elzinga analyzed 222 DAFs that Oracleprovided to the DOJ. Elzinga only ana-lyzed the ‘‘forms that [1] had U.S. custom-ers, [2] pertained to HRM or FMS soft-ware, * * * [3] [had a] net transactionprice [of] over $500,000 and [4] where thejustification section listed the competitor[or alternative solution] that was drivingthe request for the discount.’’ Tr. at2174:14–18 (Elzinga). These criteria de-creased the number of DAFs available foranalysis to just over 200. Tr. at 2175:25(Elzinga). After analyzing all the justifica-tions that were proffered by Oracle sales-persons, Elzinga created a graph thatshowed the number of times each competi-tor or alternative solution forced an Oracle


salesperson to request a discount (i.e., thenumber of times each competitor or alter-native constrained Oracle’s pricing of FMSor HRM). See Ex. P3175.

Based upon the graph, Oracle salesper-sons cited as primary justification, compe-tition from: Peoplesoft, 122 times; SAP,81 times; Lawson, 16 times; and Micro-soft and AMS, each less than 10 times.Pls Post Brief (Doc. # 366) at 12; Tr. at2177:10–11 (Elzinga). Elzinga concludedthat this discount tabulation is ‘‘very pow-erful, robust evidence’’ that the relevantproduct market is high function FMS andHRM. Tr. at 2179:7–8 (Elzinga). ‘‘I thinkthat [high function FMS and HRM] is [therelevant market] because I don’t see alter-natives outside * * * of that market, suchas mid-market, or [incumbent] or outsourc-ing, disciplining the Oracle pricing the waythe [other] two manufacturers of highfunction FMS software and HRM softwaredo, and that’s SAP and PeopleSoft.’’ Tr.at 2179:9–14 (Elzinga).

Market research studies. Independentmarket research organizations study cer-tain product markets and summarize find-

ings about any number of relevant aspectsof that market. Most of these researchorganizations conduct research and issuereports for ‘‘people who buy [the product]and want to implement it, * * * but [thesefirms are] not writing to an antitrust econ-omist or antitrust lawyer audience.’’ Tr.at 2182:12–15 (Elzinga). Elzinga foundone such market research report, conduct-ed by the Gartner Research firm, whichanalyzed the HRM pillar of the softwareindustry. Tr. at 2181: 17–18 (Elzinga).In the Gartner report, Gartner had enu-merated two characteristics upon which itchose to analyze HRM vendors: (1) ‘‘com-pleteness of vision’’ and (2) ‘‘ability to exe-cute.’’ Completeness of vision apparentlyrefers to a vendor’s level of desire to havesoftware capable of either broad and com-plex transactions (deemed ‘‘visionaries’’) orlimited and ordinary transactions (deemed‘‘niche players’’). Tr. at 2182: 17–19 (El-zinga). Ability to execute apparently refersto whether Gartner believed each vendorhad the ability to execute its HRM capabil-ity ‘‘vision’’ (e.g., high levels of functionali-ty and scalability). Tr. at 2182:24–25 (El-zinga).

According to Elzinga, the Gartner re-search only identified three firms as ‘‘vi-

sionaries’’ with a high ability to execute—Oracle, PeopleSoft and SAP. Tr. at 2183:1–


4 (Elzinga). Elzinga concluded that theGartner report is again ‘‘consistent withthe notion that there is something differ-ent about high function enterprise soft-ware from other alternatives * * * andwhen it comes to high function software,there is something different about Oracle,PeopleSoft and SAP.’’ Tr. at 2183: 16–20(Elzinga) (emphasis added).

Customers and consulting firms. Thenext strain of evidence Elzinga relied uponwere declarations of ERP customers andthe ‘‘Big Five consulting firms’’ that plain-tiffs furnished him. Tr. at 2184:8–15 (El-zinga). In particular, Elzinga pointed todeclarations of Perry Keating of BearingPoint (who also testified) and Deloitte’s[David] Dortenzo. See Elzinga demo # 6.Tr. at 2185:1–2188:9 (Elzinga). In seekingcost-effective solutions and recommenda-tions in choosing ERP vendors, many con-sumers employ consulting firms to advisethem in their negotiations with the ven-dors. Accenture, IBM Global Services,BearingPoint, Deloitte and CGEY are theconsulting firms collectively known as the‘‘Big Five.’’ At trial, plaintiffs offered thestatement of BearingPoint’s Senior VicePresident, Perry Keating. Tr. at 912:15–916:7 (Keating). Keating stated that Ora-cle, SAP and PeopleSoft ‘‘are the only[vendors] that provide a product that willbe acceptable to a large company in termsof product capabilities * * *.’’ Id.

Elzinga described the similar testimonyof Keating and Dortenzo in declarations asindicating that these Big Five systems in-tegrators most frequently recommendPeopleSoft, Oracle and SAP for ERP im-plementations. Tr. at 2186: 7–2188:9 (El-zinga).

The customers’ declarations, Elzingaconcluded, ‘‘were consistent with the hy-pothesis that there’s a distinction betweenhigh function enterprise software and themid-market * * *. [Mid-market solu-tions] are not substitutes that a hypotheti-

cal monopol[ist] * * * would be con-strained [by] in its pricing discretion [ofhigh function FMS and HRM].’’ Tr. at2184:17–22 (Elzinga).

Internal documents from ERP vendors.Elzinga was also privy to internal companydocuments, some of which he claimed wereinformative. Tr. at 2189:23–25, 2190:15(Elzinga). First, Elzinga was privy to cus-tomer surveys that had been completed byOracle ERP customers. These surveyshad been given to Oracle customers whowere classified by Oracle as having over $2billion in sales. Tr. at 2189:5–7 (Elzinga).These 28 surveys asked the Oracle custom-er to identify any other ‘‘vendors [otherthan Oracle] that were considered.’’ Id. at3–4 (Elzinga). Elzinga summarized thesurveys and concluded that PeopleSoft wasconsidered 50 percent of the time by Ora-cle customers. Ex. P3176. SAP was con-sidered 28 percent of the time and Lawsonwas considered 18 percent of the time. Id.Microsoft was considered only 4 percent ofthe time. Id.

Also of interest to Elzinga was an inter-nal document produced by Microsoft inresponse to the government’s civil investi-gative demand (CID), MS–OPCID 1610.The document was labeled ‘‘MicrosoftBusiness Solutions: Scorecard Review.’’See Elzinga demo # 8. In the document,Microsoft is characterized as worriedabout ‘‘Oracle, Peoplesoft, [and] SAP ag-gressively moving down-market, increasingpricing pressure (discounting levels) andcreating new channel programs.’’ Tr. at2192:8–11 (Elzinga). Elzinga concludedthat this document showed that Microsoft(1) recognizes a difference between mid-market and high function software and (2)does not consider itself to be in the marketfor high function ERP. Tr. at 2192:13–20(Elzinga).

From the foregoing, Elzinga crafted ametric to measure the product market.


Elzinga’s data were calculated exclusive-ly for use in this trial. In estimating theproduct market from the non-public salesdata of Oracle and PeopleSoft and thirdparty vendors obtained through the gov-ernment’s compulsory process, Pls Fact(Doc. # 356) 6.2, Elzinga applied a mini-mum threshold purchase ‘‘of $500,000 percustomer’’ to identify high function FMSand HRM. Tr. at 2210:2–4 (Elzinga). El-zinga used this threshold amount to filterout mid-market and point solution sales.Tr. at 2210:4–6 (Elzinga). Accordingly,any sale of FMS or HRM that resulted inat least $500,000 in net license revenues tothe vendor, Elzinga considered to be a saleof high function FMS or HRM and thuswas in the relevant market. Becauseplaintiffs’ product market definition has nowidely accepted meaning in the industry,there were no generally available data ex-plicating the proposed market’s partici-pants and their relevant shares to backupElzinga’s estimates.

From his numbers, Elzinga calculatedthe following United States high functionFMS market shares: SAP, 39 percent;PeopleSoft, 31 percent; and Oracle, 17percent. Tr. at 2212:22–24 (Elzinga). Amerged Oracle/PeopleSoft would, in Elzin-ga’s view, possess a 48 percent marketshare. Tr. at 2212:24–25 (Elzinga). Usingthe same data, Elzinga calculated theHHI1 in the high function FMS market tobe 2800. Tr. at 2214:17–18 (Elzinga).Based upon Elzinga’s calculations, a merg-er between Oracle and PeopleSoft wouldincrease the high function FMS HHI2 to3800. Tr. at 2214:20–21 (Elzinga).

For high function HRM, Elzinga calcu-lated PeopleSoft’s market share at 50 pe-cent, SAP at 30 percent and Oracle at 18percent; hence, in Elzinga’s view, amerged Oracle/PeopleSoft would have amarket share approaching 70 percent. Tr.at 2218:18–23 (Elzinga). Elzinga calculat-ed an HHI1 of 2800 in the high function

HRM market. Tr. at 2219:7–9 (Elzinga).Post-merger, the HHI2 would increase to5700. Tr. at 2219:10–11 (Elzinga).

Plainly, the levels of concentration re-flected in Elzinga’s testimony exceed thethresholds for ‘‘significant competitive con-cerns’’ under the Guidelines. Guidelines§ 1.51(c). Both HHI2 amounts exceed1800, and both 6HHI amounts exceed 50points. Likewise, of course, post-mergermarket shares of this magnitude wouldsatisfy the conditions to raise the anticom-petitive presumption described by the Su-preme Court in Philadelphia Nat Bank.

But for reasons explained more fullyfollowing the discussion of Oracle’s expertwitnesses, the court finds that Elzingafailed to carry the plaintiffs’ burden of (1)establishing an articulable product marketand (2) providing post-merger marketshare and HHI measurements, in a prop-erly defined market, invoking an anticom-petitive presumption under PhiladelphiaNat Bank or the Guidelines.

Oracle’s Critique of Plaintiffs’ ProductMarket Definition

Oracle painted a quite different picture.Oracle assailed plaintiffs’ high functionsoftware ‘‘label,’’ arguing that there is ‘‘nota sufficient break in the chain of FMS andHRM substitutes to warrant calling ‘high-function’ software—meaning SAP, Oracleand PeopleSoft [FMS and HRM] prod-ucts—a market unto themselves.’’ Def.Post Brief (Doc. # 365) at 17. Oracle ar-gued that the relevant product market is,at least, the entire continuum of FMS andHRM software, including those sold by so-called mid-market vendors. Id. In supportof this position, Oracle presented severalwitnesses.

Systems Integrator Witness

Oracle called Christy Bass, Global Man-aging Partner of Global Business Solutionsfor Accenture, to rebut plaintiffs’ product


market definition as well as rebut the no-tion of localized competition between Ora-cle and PeopleSoft. Accenture is the larg-est systems integrator in the world, withannual revenue exceeding $11.4 billion.Tr. at 1610:15 (Bass). Bass testified that‘‘all’’ of Accenture’s clients have high func-tion needs. Tr. at 1613:6–7 (Bass). Basstestified that several high function clients,such as Best Buy and BellSouth, had cho-sen to outsource their entire HR function.Tr. at 1648:14–19 (Bass). While some ofthese outsourcing clients were on a ‘‘one-to-one’’ outsourcing model, in which it tooka license directly from an ERP vendor,such as Oracle, Bass stated that Accentureis planning to launch the ‘‘one-to-many’’model. Tr. at 1649:14–1650:13 (Bass).Under this model, the license will be be-tween Accenture and the ERP vendor,with no contractual arrangement betweenthe customer and the vendor. Tr. at1650:3–13 (Bass). Moreover, Bass testi-fied Accenture plans to begin outsourcingFMS on a ‘‘one-to-many’’ model within thenext two years. Tr. at 1655:6 (Bass).

Bass also testified about best of breedsolutions and their potential to constrainhigh function ERP prices. Bass statedthat it was ‘‘extremely common’’ for highfunction clients to pursue a best of breedapproach. Tr. at 1668:17 (Bass). Bassstated that these best of breed solutionscould possibly offer greater functionalitythan Oracle, SAP or PeopleSoft. Tr. at1668:24–1669:3 (Bass). She also statedthat best of breed solutions put competi-tive pressure on these ERP vendors. Tr.at 1669:19–22 (Bass).

Bass rebutted plaintiffs’ assertions thatSAP was a ‘‘struggling’’ firm and alsoplaintiffs’ evidence regarding localizedcompetition between Oracle and People-Soft. Bass characterized SAP’s position inthe United States ERP marketplace as‘‘strong.’’ Further, she testified that sheconsidered SAP to have a ‘‘stronger’’ posi-

tion than either Oracle or PeopleSoft inregards to Global 2000 clients. Tr. at1621:18–23 (Bass). When asked aboutSAP’s complete exclusion from the UnitedStates banking industry, Bass concededthat such a situation existed, but opinedthat change was on the horizon. Bassdisclosed that SAP and Accenture haveentered into a ‘‘strategic alliance’’ to co-develop a banking solution for Europeanand United States banking firms. Tr. at1633:4–7 (Bass); Ex. D5001. Bass statedthat Accenture has relationships with alltwenty of the largest United States banks,and Accenture ‘‘leverag[ed] the experiencethat [Accenture] has had in the bankingindustry’’ in order to get some of thesebanks to discuss implementing the co-de-veloped software. Tr. at 1634:15–16,1635:3–10, 1636:1–6 (Bass).

The court finds Bass’ testimony to bereliable and informative on the issues ofoutsourcing and localized competition.Regarding high function clients that havechosen outsourcing as an ERP alternative,Bass gave specific examples of companies,both of which would seem to meet plain-tiffs’ high function definition, that had cho-sen to outsource their entire HRM needs.Bass’ testimony of a lack of localized com-petition between Oracle and PeopleSoftwas likewise supported by her explanationof the SAP/Accenture co-development alli-ance, under which Bass explicitly statedthat Accenture will use its leverage andexperience with United States bankingfirms in order to help SAP gain a largercompetitive share in that vertical.

Industry Witnesses: Lawson and SAP

Jay Coughlan, CEO and President ofLawson Software testified regarding hisview of the plaintiffs’ proposed productmarket and its relation to Lawson. Tr. at3586:1–13 (Coughlan). Oracle wasted notime in questioning Coughlan about plain-tiffs’ characterization of Lawson as only a


mid-market vendor. Tr. at 3596:5–9(Coughlan). Coughlan stated that he disa-greed with this view, testifying that Law-son has customers that exceed $1 billion inrevenues, employ more than 10,000 peopleand are listed among the Fortune 1000.Tr. at 3596:6–19 (Coughlan). Coughlanstated that the plaintiffs’ perception ofLawson may have been appropriate before1996, when Lawson made a conscious deci-sion to focus on specific verticals and win-ning larger shares in those verticals. Tr.at 3597:1–8 (Coughlan). The first verticalthat Lawson focused on was healthcareand today it is providing procurement andHRM for HCA, the largest health careprovider in the world with annual revenuesexceeding $20 billion. Tr. at 3600:1–4(Coughlan). Coughlan also stated thatLawson provides FMS and procurement tothe Mayo Clinic, an account for whichLawson beat both Oracle and PeopleSoft.Tr. at 3601:2–6 (Coughlan).

Coughlan stated that Lawson next fo-cused upon the retail vertical and has metwith much success. Today, Lawson pro-vides FMS to Safeway, the third largestgrocery chain in the United States withapproximate revenues of $30 billion. Tr.at 3604:1–8 (Coughlan). Lawson providesFMS to Walgreens, a convenience storechain with more than $30 billion in reve-nues. Tr. at 3604:12–21 (Coughlan).Lawson provides FMS for Target, a de-partment store chain with more than 300,-000 employees and $50 billion in revenues.Tr. at 3605:2–13 (Coughlan). The same istrue for Williams–Sonoma. Tr. at3606:16–19 (Coughlan). In the apparelarea, Lawson provides HRM and FMS toRalph Lauren and Gucci. Tr. at 3605:19–25 (Coughlan). Lawson provides HRM toMcDonald’s, a food retailer with over 100,-000 employees. Tr. at 3607:4–19 (Cough-lan).

In the public sector vertical, Lawson haswon major accounts with school districts in

Florida, Virginia and Maryland, all ofwhich Lawson competed for, and won,against Oracle and PeopleSoft. Lawsonprovides HRM to the States of Michiganand Arizona. Tr. at 3615:4–15 (Coughlan).Lawson provides HRM for the City ofDallas and the University of Wisconsin.Tr. at 3613:4–12 (Coughlan). Coughlan’stestimony continued to describe Lawsonaccounts in insurance and financial ser-vices verticals as well as individual custom-ers including Johnson & Johnson (HRM)and Sara Lee and McGraw–Hill (HRM andFMS). Tr. at 3636–3640 (Coughlan). Seealso Ex. D7140.

Moreover, Coughlan testified that Law-son software systems run in English,French and Spanish. Tr. at 3645:13–17(Coughlan). Coughlan testified that Law-son software is able to handle multiplecurrencies as well, citing one Lawson cus-tomer, Schlumberger, a major supplier tothe oil industry with $10 billion in reve-nues, 10,000 employees and internationaloperations. Tr. at 3641:23–3642:11(Coughlan). Schlumberger is utilizingLawson FMS in close to 100 countries, butnot the United States, thus showing thatthe Lawson software can handle currenciesbeyond the United States dollar. Tr. at3642:16–3643:9 (Coughlan).

Finally, Oracle asked Coughlan aboutProfessor Elzinga’s data expounding themarket shares for high function HRM andFMS. Tr. at 3648:16–3655:19 (Coughlan).Oracle offered into evidence the DOJ sub-poena to which Lawson had responded bytelling the DOJ of a large number of HRMand FMS shipments that had been madein late 2002 and throughout 2003. Ex.D7079R. This list included FMS sales toDollar Tree Store, Louisiana Pacific Cor-poration and ManuLife, with each sale to-taling more than $500,000. Ex. D7079R;Tr. at 3650:3–8 (Coughlan). Moreover,FMS suites were sent to Schlumberger,


Sara Lee and Johnson & Johnson, witheach spending more than $1 million on thesuites. Id. Accordingly, Coughlan statedthat he was perplexed when told that ac-cording to Professor Elzinga’s statistics,Lawson had no market share of the highfunction FMS market because Lawson hadmade no sales of FMS over $500,000. Tr.at 3653:10–13 (Coughlan). Further,Coughlan was told that Elzinga’s HRMdata listed Lawson as having made onlyone sale above $500,000, a sale for$995,000, leading Elzinga to call Lawson a‘‘fringe player’’ in the HRM high functionmarket. Tr. at 2219:16 (Elzinga). In re-sponse, Coughlan stated that he disagreedwith Elzinga’s calculations, citing thatLawson ‘‘had one deal alone in HRMS in[2003] that was more than one million dol-lars.’’ Tr. at 3654:1–2 (Coughlan).Coughlan stated that he disagreed withplaintiffs’ attempts to characterize Lawsonas ‘‘not a serious player [in the high func-tion market].’’ Tr. at 3655:15–19 (Cough-lan).

On cross-examination, the plaintiffs wereable to delve more deeply into the custom-er relationships that Lawson has with sev-eral of the customers discussed on direct.The City of Dallas had extensive problemswith Lawson’s HRM software, Coughlanadmitted, and its ability to handle overtimepayroll functionality, leading Dallas towithhold payments to Lawson. Tr. at3699:15–19 (Coughlan). Coughlan claimedthe problem had been corrected. Tr. at3700:11 (Coughlan). Next, an internalLawson memo showed that McGraw–Hillwas exploring the option of replacing Law-son, as was another client, Cendant. Ex.P3297. Moreover, the document statedthat Johnson & Johnson was ‘‘not purchas-ing much in the way of additional applica-tions.’’ Id. In summation, the documentseemed to call into question Lawson’s abil-ity to meet the HR needs of global organi-zations. Id. Coughlan conceded that theMayo Clinic has had problems with its

Lawson FMS software. Tr. at 3715:10–23(Coughlan). Plaintiffs went through a ser-ies of Lawson customers that have hadsome implementation or service problemwith Lawson software. Tr. at 3699–3711(Coughlan).

This evidence was elicited in an attemptto show that Lawson is not a player in thehigh function ERP market. The evidencedid show the existence of implementationor service problems. But the customersall appeared to fit plaintiffs’ definition ofhigh function customers. Hence, this lineof inquiry did not appear to demonstrateLawson’s absence from this or any suchmarket, only that some Lawson customershave had problems with its software. Thecourt, therefore, discounts Coughlan’scross-examination testimony for the pur-pose for which it was apparently offered.Plaintiffs did not show that implementationor service problems were absent or lessfrequent in Oracle, PeopleSoft or SAPproducts. Accordingly, the court creditsCoughlan’s testimony regarding large andcomplex customers that have chosen Law-son ERP to meet their FMS and HRMneeds. Not only was this evidence uncon-tradicted, but the testimony was amplysupported by many exhibits.

Richard Knowles, Vice President of Op-erations for SAP America, was called byOracle to refute the plaintiffs’ productmarket definition as well as to poke holesin plaintiffs’ theory of unilateral anticom-petitive effects. Tr. at 2805:4–9 (Knowles).At the outset, Knowles clarified some ofthe terms used in this case, or at least asthose terms are understood by SAP. ‘‘Highfunction’’ has no meaning apparently.SAP looks to customer characteristics indetermining whether a vendor is mid-mar-ket or high function. SAP considers acustomer to be mid-market if it has reve-nues less than $1.5 billion, but more than$200 million. Tr. at 2818:9–19 (Knowles).


A customer above $1.5 billion is considereda ‘‘large enterprise.’’ Tr. at 2819:14–15(Knowles). But Knowles stated that char-acterizing a customer as one or the otherwas far from ‘‘an exact science.’’ Tr. at2820:18–19 (Knowles).

Oracle then proceeded to ‘‘name drop’’ alarge number of SAP clients: Deloitte &Touche, Accenture, Halliburton, MCI,SBC, T–Mobile, AOL, Starbucks, Nike,Home Depot and Barnes & Noble, allclearly up-market customers. Tr. at2829:6–2831:19 (Knowles). This evidencetended to rebut the suggestion that SAPwas a struggling firm with substantial dis-advantages in the United States. Next,Oracle questioned Knowles regarding twospecific examples in which SAP had com-peted head to head against Oracle andother ERP vendors. First, Oracle pre-sented an SAP DAF regarding a proposedERP license transaction with ExpressJet,a company with approximately $1.5 billionin revenues, thus making it a large enter-prise. D5641R at 1; Tr. at 2839:23(Knowles). As with the DAFs used byOracle, the SAP DAFs had a column fordenoting the competitor that was requiringor motivating the increased discount re-quest. D5641R at 1. In the case withExpressJet, SAP was originally competingagainst PeopleSoft, Lawson, Exact, Micro-soft, Oracle and Ultimate. Ex. D5641R at2. Knowles stated that he recognized thename Lawson and that SAP ‘‘of course’’competes with Lawson. Tr. at 2841:8–12(Knowles). Moreover, once ExpressJethad narrowed the six vendors down tothree, it was a contest with Lawson, Oracleand SAP. Tr. at 2842:23–2843:5 (Knowles).

Knowles stated that SAP was ‘‘agnostic’’about which competitor makes it to thefinal round, because SAP is going to givethe same level of discount regardless ofthe competitor. Tr. at 2848:7–10(Knowles). Oracle then introduced anoth-er SAP DAF, this time for Kellogg, Brown

& Root, a subsidiary of Halliburton. Ex.D5649R at 1. This form listed the justifica-tion for the discount as the ‘‘extreme com-petition’’ between Oracle and SAP. Id.Knowles stated that this type of scenariowas to be expected, as SAP views Oracleto be ‘‘highly aggressive’’ on pricing. Tr.at 2856:10–11 (Knowles).

Next, Oracle introduced an email fromBill McDonald, the CEO of SAP America.Ex. D5636. The email contained Micro-soft’s second quarter earnings for 2004.Id. at 1. The document began by reading:‘‘These guys are here!’’. Id. Knowles stat-ed that McDermott was referring to Mi-crosoft’s 32 percent year-over-year in-crease in the EAS market. Tr. at 2892:4–23 (Knowles).

Finally, Oracle questioned Knowlesabout any apprehensions SAP felt regard-ing increased prices should the proposedmerger of Oracle and PeopleSoft be con-summated. Tr. at 2858:9–11 (Knowles).Knowles responded that SAP has a neutralopinion on the merger. Knowles statedhis belief that the merger will actuallymake the ERP market more competitive.Tr. at 2858:20–21 (Knowles).

On cross, Knowles conceded that thereason SAP America exists is because cus-tomers in the United States ‘‘want to havesomebody here present to deal with inbuying the type of software that [SAP]sells.’’ Tr. at 2902:12–15 (Knowles). Next,Knowles stated that SAP views Lawson asa ‘‘mid-market company.’’ Tr. at 2924:24(Knowles). This characterization appearsto rest on SAP’s labeling as mid-market ofcustomers with less than $1.5 billion inrevenues, a substantially different demar-cation from plaintiffs’ labeling of a mid-market customer as one that does not buysoftware packages exceeding $500,000.See Tr. at 2924:5–7 (Knowles). According-ly, the court accords no weight to Knowles’statement inasmuch as it was offered to


show that Lawson does not compete in thehigh function market. Otherwise, thecourt finds Knowles’ testimony to be reli-able and uncontradicted.

Outsourcing Witnesses

Michael Sternklar, Executive Vice Presi-dent of Fidelity Human Resources Ser-vices Company, testified regarding Fideli-ty’s outsourcing solutions for HR needs.Tr. at 3124–3126 (Sternklar). Sternklarstated that Fidelity currently has a licensewith Oracle for HRM software. Tr. at3130:2–7, 3135:3–8 (Sternklar). Sternklarlisted several of Fidelity’s ‘‘large’’ custom-ers: Bank of America, IBM, AmericanCorporation and Asea Brown Boveri(ABB). Tr. at 3136:13–3137:8 (Sternklar).Sternklar described the procurement pro-cess by which Fidelity won the ABB ac-count. ABB made its choice between buy-ing an in-house system from PeopleSoft orSAP, or instead, outsourcing ABB’s HRMneeds through Fidelity. Tr. at 3138:14–25,3139:10–13 (Sternklar). ABB chose Fideli-ty over PeopleSoft and SAP. Tr. at3140:17–18 (Sternklar). Sternklar statedseveral reasons why a customer wouldchoose outsourcing over an in-house ERPsystem. One important reason, Sternklarstated, was the ‘‘continued investment’’ in-volved in buying an in-house ERP systembased upon the need continuously to up-grade such a system. Tr. at 3139:23–25(Sternklar).

Jay Rising, President of National Ac-counts at ADP also testified about what hecalled ‘‘upgrade treadmill’’ and ‘‘hiddencosts’’ that are involved in package soft-ware. Tr. at 4094:15–22 (Rising). BothSternklar and Rising testified that thereare no such costs associated with outsourc-ing because the outsourcer itself, not thecustomer, handles all upgrades and main-tenance. Tr. at 3140:2–5 (Sternklar);4093: 12–16 (Rising). The client need notbother with such hassles.

Both witnesses also testified that out-sourcing companies are able to handle theHR needs of companies with large num-bers of employees. ADP has 1000 custom-ers that have over 1000 employees. Tr. at4097:21–25 (Rising). ADP’s client list in-cludes Comcast, Sysco, Xerox and Tyco.Tr. at 4100:13–24 (Rising). Fidelity out-sources for Bank of America which cur-rently has between 170,000 and 180,000employees. Tr. at 3145:18–25 (Sternklar).

Finally, Sternklar stated that Fidelitywas currently in the process of creating itsown software called HR Access. Tr. at3152:3–3153:13 (Sternklar). Fidelity’s goalis to move all customers onto HR Accesswithin the next two years and cease usingOracle software completely. Tr. at3154:9–15 (Sternklar).

The evidence of both of these outsourc-ing witnesses was reliable and amplysupported by specific examples of highfunction customers that had chosen tooutsource with Fidelity or ADP as anERP alternative. Accordingly, the courtcredits this testimony in determiningwhether outsourcing solutions have aprice-constraining effect on ERP vendors.

Expert Witnesses: Hausmanand Campbell

Oracle did not propose a product marketdefinition. Instead, Oracle picked apartplaintiffs’ market definition piece by piece.Two expert witnesses, Professor JerryHausman, an industrial organization econ-omist at MIT, and Tom Campbell, dean ofthe Haas Graduate School of Business atthe University of California (Berkeley) tes-tified for Oracle. Among other importantpositions in government, Campbell servedas director, Bureau of Competition, at theFTC. Both Hausman and Campbell as-sailed plaintiffs’ product market definition,describing it as vague, unrealistic and un-derinclusive. As with Elzinga, the courtfinds both Hausman and Campbell to be


well qualified to offer their opinion testi-mony.

Vague. Hausman characterized the‘‘high function’’ label as vague and too‘‘hard to get your arms around.’’ Tr. at3807:14–15 (Hausman). He cited plain-tiffs’ changing description of ‘‘high func-tion’’ ERP as illustrating the unreality ofplaintiffs’ proposed product market defini-tion. Tr. at 3809:20–3810:9 (Hausman).At first, plaintiffs argued for a customer-based product definition. Campbell char-acterized this initial customer-based mar-ket definition as ‘‘unprecedented’’ and‘‘unusual.’’ Tr. at 2704:6 (Campbell).Hausman asserted that plaintiffs, inreaching this strange product market,clearly worked backwards from their de-sired result: finding a group of customersall of which had purchased SAP, Oracle orPeopleSoft ERP, then claiming that thosecustomers were ‘‘similarly situated’’ anddefined the market. But, at trial, Haus-man noted, plaintiffs shifted ground andargued that the high function market wasbased upon ‘‘product characteristics’’ ofthe software, such as functionality andscalability, not the customers who buy it.Tr. at 3809:20–3810:3 (Hausman). Seealso Pls Post Brief (Doc. # 366) at 3–4.

Even accepting the plaintiffs’ secondversion of high function software, both ex-perts asserted that the term is too impre-cise to define a market. Hausman con-tended that Elzinga himself admitted thatthe high function definition contained no‘‘quantitative metrics’’ that could be usedto distinguish a vendor of high functionERP from a vendor of mid-market soft-ware. Tr. at 3807:16–17 (Hausman). SeeTr. 2151:18–2152:3 (Elzinga). Hausman il-lustrated his point by reference to Manu-Life Insurance Company, the fifth largestinsurance company in the United Stateswith offices throughout North America.ManuLife has complex needs and transac-tions and thus by any objective measure

would fit in plaintiffs’ high function mar-ket. But plaintiffs, for a reason Hausmansaid plaintiffs left unexplained, consideredManuLife to be a mid-market purchaserand therefore excluded from the plaintiffs’market definition. Tr. at 3840:17–3841:13(Hausman). The same applies to Johnson& Johnson and Safeway, both consideredby plaintiffs as mid-market customers be-cause they bought ERP solutions fromvendors that Elzinga and plaintiffs put inthe mid-market. But plainly these firmsfit plaintiffs’ description of enterpriseshaving high functional needs. So, conclud-ed Hausman, plaintiffs have provided noobjective way to distinguish ERP licensesin the high function market from those inthe mid-market.

Both Hausman and Campbell made theobvious point that if the market is notprecisely defined, then the market partici-pants and their relative shares will be‘‘economically inaccurate.’’ Tr. at 2702:16–19 (Campbell); 3793:9–11 (Hausman). Re-ferring to plaintiffs’ customer witnesses,Hausman asserted that surveys that askcustomers what their preferences are orwhat their hypothetical actions ‘‘would be’’are known to be unreliable and subjective.Id.

Oracle summarized Hausman’s vague-ness argument by claiming ‘‘there must bea clear break in the chain of substitutes inorder for separate markets to be found.’’Def Post Brief (Doc. # 365) at 17. Accord-ing to Oracle, ‘‘[T]here is clearly not asufficient break in the chain of FMS andHRM substitutes to warrant calling * * *software [sold by] Oracle, PeopleSoft andSAP, a market unto themselves.’’ Id. Ifsuch a clear break exists, plaintiffs havenot proven it by a preponderance of theevidence, Oracle argued in closing. Id.

Disconnected. Oracle also argued thatplaintiffs’ product market definition ‘‘doesnot address the market reality’’ of the way


software is sold, a point to which Hausmantestified. Def. Post Brief (Doc. # 365) at1. Hausman posited that FMS and HRMare not products in and of themselves.Rather, ‘‘90 percent’’ of companies ‘‘arebuying more than just FMS, more thanjust HRM. * * * [they are] buying bun-dles of software.’’ Tr. at 3815:10–12,3813:12–22 (Hausman).

Hausman gave as an example a consum-er purchasing a single package of softwarefrom PeopleSoft that included FMS, HRM,EPM and CRM pillars. In such a bundle,PeopleSoft would not offer discounts basedon the individual pillars. Rather, People-Soft would give a ‘‘blended discount’’across all products in the bundle in orderto ensure that the consumer buys all thepillars from PeopleSoft. Tr. at 3814:3–22(Hausman). If the vendor does not offeran acceptable discount, then the consumercan threaten to buy one of the pillars, suchas CRM, from a best of breed vendor suchas Siebel. Tr. at 3815:1–6 (Hausman).Based upon this analysis, Hausman opinedthat the presence of best of breed vendorsconstrains the prices that the ERP ven-dors can charge for a bundle of software.Tr. at 3814:18–22 (Hausman).

Underinclusive. Finally, Oracle’s wit-nesses stated that even if one assumesthat a ‘‘high function HRM and FMS’’market does exist and the market can bedemarcated from other solutions, there areviable substitutes for high function ERPthat must be included in the product mar-ket. Specifically, Oracle argued that (1)mid-market vendors, (2) outsourcing, (3)incumbent systems, and (4) best of breedsolutions, discussed above, must all be in-cluded in the product market, as all arepotential substitutes constraining a post-merger SSNIP.

Incumbent systems, also called legacysystems, refer to the FMS and HRM soft-ware systems that the DOJ’s ‘‘enterprisecustomers’’ already have in operation.

These are the systems that the new soft-ware from Oracle or PeopleSoft or SAPwill replace, should a consumer choose topurchase an integrated suite from one ofthe high function vendors. Oracle arguedthat if a post-merger Oracle/PeopleSoftimposed a SSNIP, consumers could con-strain that SSNIP by simply refusing tobuy high function FMS and HRM andchoosing to use already existing products.Tr. at 3821:1–9 (Hausman). Hausmanstated that the cost of maintaining andupgrading incumbent systems has been de-creasing recently so that these systemshave become a ‘‘credible threat’’ to ERPvendors. Tr. at 3821:13 (Hausman). Ac-cordingly, if a customer finds a post-merg-er price offer too high, it can almost al-ways credibly claim it will not buy theproduct and instead continue to operate itsincumbent system. Tr. at 3821:13–14(Hausman).

Campbell stated that ‘‘20 to 30 percentof the time, even after negotiations havestarted, the purchaser will opt to drop out’’and remain with the system it already has.Tr. at 2708:23–25 (Campbell). Campbellclaimed that this factor must be taken intoaccount when calculating market shares,otherwise ‘‘you’ve made a very serious mis-take in calculating your market shares,’’because 20 to 30 percent of the relevantcustomers’ actual behavior is being ig-nored. Tr. at 2709:1–6 (Campbell).

Regarding outsourcing, Hausman pre-sented evidence of over twenty large en-terprises, such as Bank of America and AT & T, who currently outsource all orsome of their HRM needs. Tr. at 3825:19–25 (Hausman). And this phenomenon wasoccurring long before Oracle made itstake-over offer to PeopleSoft. These largeenterprise customers would not be out-sourcing if they did not find this option tobe equal to or better than the purchase ofhigh function software from a vendor. Tr.


at 3828:19–23 (Hausman). If this manycorporations can currently have theirHRM needs effectively met by outsourc-ing, it only follows that many more cus-tomers could follow suit should a post-merger SSNIP occur in the high functionmarket. Tr. at 3829:1–3 (Hausman).

Hausman gave the example of MIT, hisemployer, outsourcing its HRM to Fideli-ty, who he claims do ‘‘a heck of a lotbetter’’ than MIT personnel. Tr. at 3825:4(Hausman). Hausman presented evidencethat many companies have chosen out-sourcing; these include: Bank of America,Motorola, International Paper, McKesson,American Express and Sony. Tr. at3829:21–23 (Hausman). These are ‘‘so-phisticated’’ companies, with a lot of com-plex transactions, and they have clearlyfound outsourcing a satisfactory alterna-tive. Id. Hausman’s demonstratives alonelisted seven outsourcing firms capable ofhandling the HR for large companies;these include Fidelity, Accenture, ACS,Exult and Mellon, among others. Haus-man demo # 10.

Accordingly, both Campbell and Haus-man asserted that any product marketmust include outsourcing solutions as aviable substitute to which consumers canturn in the event that a merged Ora-cle/PeopleSoft imposes a SSNIP.

Finally, Oracle attempted to show thatthe products of so-called mid-market ven-dors, such as Lawson and AMS are rea-sonably interchangeable for those of thealleged high function vendors, Oracle, Peo-pleSoft and SAP. Accordingly, Hausmanstated that any market definition that isdevoid of these vendors is too narrow. Tr.at 3939–3940 (Hausman). Hausman pre-sented evidence of over thirty consumers,all of which have large and complex needs,and all of which had chosen to use Lawsonor AMS for their FMS and HRM needs.Lawson’s customers include: Johnson &Johnson, Walgreens, Target, Williams–So-

noma, Jack in the Box, the Federal Re-serve Bank and Safeway. Hausman demo# 11. AMS’ customers include: UnitedStates Environmental Protection Agency,United States Postal Service, Library ofCongress, Internal Revenue Service andthe DOJ. Id. Very telling to Hausman wasthe fact that the DOJ, two weeks afterbringing this case, chose to buy AMS FMSfor $24 million, ranking AMS better thanOracle or PeopleSoft in the DOJ’s view forthe DOJ’s needs. Tr. at 3842:7–13 (Haus-man).

Hausman admitted that these vendorsare ‘‘not PeopleSoft,’’ nor do they ‘‘aspireto be.’’ Tr. at 3839:4–6 (Hausman). Healso admitted that these three ‘‘cannot cur-rently satisfy the entire market as definedby the DOJ.’’ But ‘‘you do not have to beatPeopleSoft to constrain it’’ argued Haus-man. Tr. at 3839:20–21 (Hausman). Thequestion is not whether the entire marketwould switch to these other vendors in theevent of a SSNIP, the question is whetherenough consumers could potentially turnto a product to meet their needs, therebymaking a SSNIP unprofitable. Clearly, ifthe high function needs of Johnson &Johnson and the DOJ are met by thesemid-market vendors, then many othercompanies could also do so in the wake ofa SSNIP. Accordingly, these two mid-market vendors should be included in theproduct market.

Infrastructure layer. Two of defen-dant’s expert witnesses discussed the in-frastructure layer and its impact on theproduct markets. Tr. at 4138–4145 (Kut-nik); Tr. at 4364–4369, 4397–4398 (Teece).Traditionally, ERP software containedboth business logic and applications ser-vices. Business logic is the logical struc-ture of the business process being auto-mated. Applications services are toolsthat support business logic across differentbusiness applications. Applications ser-


vices include directory services, securityfeatures and content management tools.Web services are a type of applicationsservice.

Recent innovations in software technolo-gy have led to a ‘‘decoupling’’ of businesslogic from applications services. These in-novations have resulted in the creation ofan ‘‘infrastructure layer’’ that standardizesmany of the applications services that wereonce incorporated with the business logicin an EAS program. The infrastructurelayer has also been referred to as the‘‘integration layer,’’ the ‘‘applications ser-vices’’ layer and the ‘‘composite applica-tions’’ layer. Tr. at 325–31 (Bergquist).

Infrastructure layer products and ERPsoftware share some degree of substituta-bility in that both address integration.Developments in infrastructure layer tech-nology allow greater interoperability andeasier horizontal integration. Ex. D7143(Mills 5/27/04 Dep) at Tr. 59–61; Tr. 2886–89 (Knowles); Tr. 4150:9–19 (Kutnick).Similarly, pre-integration in ERP softwaresuites allows greater interoperability andeasier horizontal integration. Because onecan choose more robust infrastructure lay-er products instead of pre-integration, theinfrastructure layer is a partial substitutefor the pre-integration in ERP softwaresuites.

Oracle’s experts Kutnick and Teece tes-tified that the emergence of the infrastruc-ture layer constitutes a paradigm shift inERP software products and affects theproper product market definition.

The following facts suggest that infra-structure layer products should not be in-cluded in the same relevant market asERP software. First, the overlap in sub-stitutability between infrastructure layerproducts and ERP software is limited.ERP software products perform a largenumber of functions that are not per-formed by infrastructure layer products,and vice versa. Accordingly, sellers of

infrastructure layer products likely couldnot constrain market power of a hypotheti-cal monopoly over ERP software.

Second, the integration offered by infra-structure layer products is a poor substi-tute for pre-integration in ERP softwaresuites. Pre-integration allows tighter inte-gration than the integration offered byinfrastructure layer products. Certainfunctions previously performed within theERP software layer are now performed inthe infrastructure layer. Infrastructurelayer products, however, do not containbusiness logic. Tr. at 4144:8–11, 4187(Kutnik); Tr. at 1813–1814 (Wilmington);Tr. at 331–332 (Bergquist). Because infra-structure layer products do not containbusiness logic, a purchaser could notchoose a more robust infrastructure layerproduct instead of ERP software. Accord-ingly, the decoupling of the infrastructurelayer from the ERP software layer doesnot suggest that the infrastructure layerproducts are partially substitutable forERP software.

Oracle’s experts Kutnick and Teece con-tend that the emergence of the infrastruc-ture layer constitutes a paradigm shift inERP software products. The age of infra-structure layer products calls into questionthis contention. See D7143 (Mills 5/27/04Dep) Tr. at 30–31 (stating that IBM’s mid-dleware products have been in the marketfor nearly twenty years); Tr. at 420 (Kut-nik) (testifying that applications servershave been available for seven to eightyears); Tr. at 3414:2–18 (Wohl) (notingthat Oracle’s applications server has beenthrough several versions); Tr. at 328(Bergquist) (testifying about the evolutionof web services protocols).

Even if the emergence of the infrastruc-ture layer will have a substantial impact onthe EAS software industry, more robustinfrastructure layer products both enhanceand diminish the likelihood of stack compe-


tition. On the one hand, decoupling appli-cations services from the business logicprovides the interoperability standard nec-essary to create multi-seller clustering.See Tr. at 4378–4379 (Teece). On theother hand, enhanced infrastructure layerproducts increase interoperability withother stacks. See Tr. at 2885–2889(Knowles); Tr. at 1637:7–22 (Bass);P3337; D7143 (Mills 5/27/04 Dep) Tr. 59–61; Tr. at 4150:9–19 (Kutnik).

Findings of Fact: ProductMarket Definition

[13] In order to sustain plaintiffs’product market definition the court mustfind, by a preponderance of the evidence,that plaintiffs’ have shown an articulableand distinct product market for HRM andFMS sold by Oracle, PeopleSoft and SAPonly that does not include mid-market soft-ware, outsourcing solutions, best of breedsolutions, legacy systems or the infrastruc-ture layer.

Based upon a review of the law and theevidence, the court concludes that theplaintiffs have not met their burden ofestablishing that the relevant product mar-ket is limited to so-called high functionFMS and HRM sold by Oracle, PeopleSoftand SAP. The equivocal and vague evi-dence presented by plaintiffs at trial doesnot permit the court to exclude mid-mar-ket vendors, outsourcing or best of breedsolutions from any product market thatincludes ERP software sold by Oracle,PeopleSoft and SAP.

For reasons discussed above, the courtcannot rely upon the testimony of the cus-tomer witnesses offered by plaintiffs indetermining if plaintiffs have met theirburden. Likewise, the testimony of allthree industry witnesses offered by plain-tiffs affords no reliable or articulable basisto distinguish a high function product mar-ket. Ironically, much of plaintiffs’ testimo-ny supports a finding of no clear or articu-lable distinction.

Accordingly, the full weight of the plain-tiffs’ product market burden fell at trialupon Elzinga. In resolving the battle ofthe expert witnesses on product definition,the court must conclude that Oracle’s wit-nesses presented the better and more con-vincing case. Elzinga for all his indubita-ble credentials as an economist seemedmostly to apply the techniques of his avo-cational interest in mystery writing. SeeEx. P4014A. The evidence Elzinga mar-shalled was circumstantial and highly qua-litative.

Elzinga’s tabulations of concentrationstatistics from responses to the DOJ CIDs,Elzinga demo. ## 10–11, suffer from sev-eral shortcomings. Elzinga defined highfunction ERP as any sale in excess of$500,000. As the DAFs establish, ERPvendors sell a cluster of products. Salesexceeding a half-million dollars, therefore,are likely in many instances, if not most, toinclude pillars other than FMS and HRM.Elzinga’s chosen demonstrative, Ex.4015A, will make the point. The sale inquestion, to Teradyne Corporation, metthe $500,000 threshold. Ex. 4015A at OR-LITE0086650. Yet the discount Oracleoffered on the HRM pillar license fee was100 percent, and the bundle included mo-dules in the SCM pillar along with modulesin the HRM pillar. Despite this, Elzingatabulated this entire transaction as an Ora-cle HRM sale, even though Oracle ap-peared to give away for free an HRMlicense in order to sell modules in the SCMpillar. Id. at ORLITE0086654. The courthas not attempted to retabulate marketshares to correct for these problems.

Elzinga’s other statistical tabulations aresketchy at best. The tabulation of Oraclecustomer surveys was a tiny sample ofonly twenty-eight sales opportunities. El-zinga demo # 7. The roster of OracleDAFs was also short. Elzinga demo # 3.But even more troubling, as pointed out in


connection with the Teradyne sale, is thatthese tabulations did not break out FMSand HRM sales from the bundles in whichthey were sold. Again, the DAFs registerprices and discounts on a mixture of differ-ent pillars and modules. Metaphorically,Elzinga did not separate the wheat fromthe chaff.

Not only does the court find Elzinga’sdata to be unreliable in establishing a dis-tinct and articulable product market, butElzinga himself admitted that plaintiffs’product market has no ‘‘quantitative met-ric’’ that could be used to determine thedistinction between a high function productand a mid-market product. Tr. at 2311:3–17 (Elzinga). Rather, Elzinga kept tellingthe court that there is ‘‘something differ-ent’’ about the products sold by Oracle,SAP and PeopleSoft. But the court can-not delineate product boundaries in multi-billion dollar merger suits based upon themere notion that there is ‘‘something dif-ferent’’ about the merging products and allothers, especially when that ‘‘somethingdifferent’’ cannot be expressed in terms tomake a judgment of the court have mean-ing. More is required.

Accordingly, based upon the evidencepresented at trial, the court concludes thatthe following products cannot be excludedfrom the relevant product market for pur-poses of analyzing the effects of this merg-er.

Outsourcing. Professor Hausman pre-sented evidence of over twenty large en-terprises that currently outsource all orsome of their HRM needs. Furthermore,the testimony of Peters, Bass, Sternklarand Rising all support Hausman’s conten-tion that large companies can, and do, havetheir HRM needs effectively met by out-sourcing. Accordingly, outsourcing solu-tions cannot be excluded.

Plaintiffs argue that because several ofthe outsourcing firms themselves use Ora-cle, SAP or PeopleSoft, these outsourcing

firms do not count as independent compet-itors. But the court finds the testimony ofBass and Sternklar regarding ‘‘blanket li-censes’’ or ‘‘one-to-many’’ licenses to bethe most reliable on how outsourcingworks. Most outsourcers that handleHRM needs for large enterprises eitherhave, or soon will have, a type of blanketcontract with an ERP vendor. Underthese contracts, the software vendorsagree to provide software to the outsour-cer at a set price up to a certain number ofemployees, or ‘‘seats,’’ usually numberingwell into the millions of employees. Fidel-ity’s contract with Oracle provides for a‘‘seat’’ capacity of 2 million employees,with Fidelity having the option to increasethe number of employees at a pre-set fee.Ex. D7158. So if Company X chooses tooutsource through Fidelity, which may beoperating on Oracle software, there is nodirect connection between Oracle andCompany X. There is no license betweenOracle and Company X and no chance forOracle to take advantage of Company Xwhich has no ‘‘post-merger’’ choice in ERPsoftware. Company X is merely more‘‘seats’’ in Fidelity’s millions of emptyseats under its blanket contract.

Moreover, several outsourcing firms cur-rently use their own proprietary software,such as Hewitt and ADP. Fidelity has alsobegun the process of migrating clientsfrom Oracle software to Fidelity’s ownsoftware. Tr. at 3154:3–15 (Sternklar).

Mid-market vendors. The court is per-plexed about plaintiffs’ position that ‘‘mid-market solutions’’ are not part of the prod-uct market for high function ERP. Plain-tiffs claim that mid-market vendors, suchas Lawson and AMS could not constrain apost-merger SSNIP. Pls. Post Brief (Doc.# 366) at 14. Such a statement clearlyimplies that plaintiffs do not view Lawsonand AMS as high function vendors. ButElzinga’s high function market share cal-


culations showed Lawson and AMS eachhad market shares. See Elzinga demo.## 10–11. Further, Elzinga stated thathis calculations probably understated Law-son’s market share in the high functionmarket. Lawson and AMS plainly cannothave market shares in the high functionmarket if they are not a part of it. Ac-cordingly, plaintiffs’ own evidence showsthat either (1) Lawson and AMS are asubstitute for high function vendors or (2)no substantive demarcation between thesetwo types of vendors exists. This evidenceglaringly shows that plaintiffs have failedto prove a distinct relevant product mar-ket for this court to analyze.

As surprising as plaintiffs’ evidence andstatistics on the mid-market is one of theplaintiff’s actual behavior. Plaintiffs char-acterize vendors that serve the mid-mar-ket as ‘‘hav[ing] limited capacity to supportcustomers with diverse operations such asmultiple geographic locations, distinct legalentities * * * or numerous lines of busi-ness.’’ FAC (Doc. # 125) ¶ 12 at 8. But,soon after filing its complaint, the UnitedStates Department of Justice itself—whichsurely meets at least two of these crite-ria—chose AMS, a so-called mid-marketvendor, to meet its HRM and FMS needs.The DOJ chose AMS over Oracle and Peo-pleSoft.

Plaintiffs’ statistics, expert witness andbehavior all treat mid-market vendorsLawson and AMS as part of the highfunction market. The court sees no rea-son why it should not follow suit.

Microsoft. As discussed above the courtfinds Burgum’s testimony regarding Mi-crosoft’s entry into the up-market to beincredible. The testimony of Keating, aswell as BearingPoint’s homepage, make itclear that Microsoft has every intention ofusing Axapta and BearingPoint to competefor so-called up-market business. Fur-thermore, Allen’s testimony about thestruggle of J D Edwards in trying to enter

the up-market does not apply to Microsoft.Microsoft has the money, the reputationand now, due to the BearingPoint alliance,it has the sales force necessary to becomea major competitor for up-market busi-ness. Accordingly, the court finds thatMicrosoft will be a viable substitute for asignificant number of consumers should apost-merger Oracle impose a SSNIP in itspricing of ERP software.

Best of breed solutions. The court doesnot dismiss defendant’s bundle argumentas an ‘‘elaborate distraction’’ or ‘‘economi-cal nonsense’’ as plaintiffs urge. Pls PostBrief (Doc. # 366) at 21–22. The reality ofthis industry is that 90 percent of consum-ers purchase software ‘‘bundles’’ contain-ing several pillars; rarely does a consumerpurchase a single pillar. Tr. at 3815:10–13(Hausman). FMS and HRM pillars typi-cally are sold in a bundle along with addi-tional kinds of EAS, such as CRM orSCM. Further, the discounts that are of-fered to potential consumers are based onthe value of the entire bundle, not simplybased upon the presence of an HRM orFMS pillar. Tr. at 3813:23–3814:1 (Haus-man). Accordingly, when Oracle or Peo-pleSoft offer a discount on a bundle, theyare doing so in order to ensure that thecustomer purchases all the pillars fromOracle or PeopleSoft, rather than turn to abest of breed vendor.

Incumbent solutions. The court, how-ever, is not persuaded that incumbent so-lutions would be able to constrain a post-merger Oracle from imposing a SSNIP.Companies can, and apparently do, threat-en to ‘‘do nothing,’’ in hopes of getting abetter price on ERP software. See Camp-bell demo. ## 20–21. But it is highlyunlikely that any monopolist would see thisthreat as ‘‘credible,’’ thereby preventing aSSNIP. Given the ever-changing condi-tions of both the regulatory and technolog-ical aspects of human resources and finan-


cial management, it is hard to sustain theidea that large corporations would ratheremploy an antiquated software systemthan pay 10 percent more for modern andcontinuously maintained products. Such achoice in today’s business world would beextremely risky and unlikely.

Accordingly, without a relevant markethaving been established, the court cannotconduct a burden-shifting statistical analy-sis under Philadelphia Nat Bank, muchless hold that plaintiffs are entitled to sucha presumption. Nor, of course, can thecourt apply the concentration methodologyof the Guidelines. See Guidelines § 1.51.

Plaintiffs’ Proposed Geographic Market

Assuming that high function FMS andHRM is the relevant product market,plaintiffs claimed that the relevant geo-graphic market is the United States. PlsPost Brief (Doc. # 366) at 22. Again,plaintiffs relied heavily on Elzinga’s testi-mony. In reaching this market definition,Elzinga ironically enough did not rely uponthe oft-used Elzinga–Hogarty (E–H) test,which he admitted has been used in ‘‘doz-ens and dozens of merger cases’’ andwhich he himself co-developed. Tr. at2154:22–23 (Elzinga).

In informal terms, the E–H test ‘‘mea-sures the accuracy of a market delineationby determining the amount of either im-ports into or exports from a tentative mar-ket. The test is based on the assumptionthat if an area has significant exports orimports, then that area is not a relevantgeographic market. Under the [test], ex-ports or imports greater than 10% suggestthat the market examined is not a relevantmarket.’’ United States v. Country LakeFoods, Inc., 754 F.Supp. 669, 672 n. 2(D.Minn.1990).

Elzinga stated that he did not believethe E–H test ‘‘fit this particular antitrustcase.’’ Tr. at 2154:25–25 (Elzinga). In-stead, Elzinga relied solely upon theGuidelines ‘‘hypothetical monopolist’’ test

in determining the geographic market.Tr. at 2204:1–11. (Elzinga). See Guide-lines § 1.21. ‘‘I am persuaded that theUnited States [is the geographic marketbecause] if [some] one were the sole sup-plier of high function FMS and HRM* * * in the US, and [he imposed aSNNIP], he would not be thwarted orundercut by economic * * * agents out-side the United States.’’ Id.

Elzinga cited several relevant factorsthat led him to believe the Guidelines re-quired a United States-only geographicmarket. Tr. at 2203:24–25 (Elzinga).

Where software code is written is notrelevant to geographic market. ‘‘The[product] market here is high functionFMS and HRM, and that is not just code.What you buy when you buy this product ** * is a relationship.’’ Tr. at 2154:10–14(Elzinga) (emphasis added). Plaintiffsurged the court to exclude from the geo-graphic market the site of manufacture.Hence, Elzinga urged the court to lookbeyond the location of manufacture forFMS and HRM. Since all of SAP’s soft-ware is manufactured in Germany andSAP indisputably produces high functionERP, inclusion of SAP’s site of manufac-ture would wholly undermine plaintiffs’proposed geographic market.

Rather, Elzinga stated that the relevantfactor in determining the geographic mar-ket is how the products are ‘‘marketed andsupported’’ (i.e., the relationship) betweenthe ERP vendor and the consumer. Tr. at2202 (Elzinga). Elzinga argued that pur-chasing high function FMS and HRM en-tails installation, implementation, mainte-nance and upgrades—a relationship thathas an inherently ‘‘local’’ aspect. Tr. at2154:21–25 (Elzinga). Accordingly, sincethe relevant factor is the marketing andsupport of the software (which occurs inthe United States) and not the ‘‘shipment’’of the software from the manufacturing


site (which could occur outside the UnitedStates), the E–H ‘‘shipments’’ test is notappropriate for this merger analysis. Tr.at 2205:11–14 (Elzinga).

Under the Guidelines, because the rele-vant factor of ‘‘relationship’’ occurs onlywithin the United States for United Statescustomers, these customers could not seeksubstitutes abroad in the event of aSSNIP, thus making the United States thegeographic market, according to Elzinga.

No arbitrage exists in this market. Ar-bitrage occurs when a consumer of a prod-uct buys the product from a vendor in onegeographic location at a low price, but thensells the product to another consumer in adifferent geographic location for a higherprice. Tr. at 2157:15–19 (Elzinga). Arbi-trage is a factor that Elzinga stated can‘‘stitch’’ together two geographic locationsinto ‘‘one geographic market’’ for mergeranalysis. Tr. at 2157:20–22 (Elzinga). El-zinga illustrated the phenomenon of arbi-trage for the court via a precious stonehypothetical. ‘‘If the price of diamondsgot relatively high in the United States,compared to * * * Europe, * * * arbi-tragers could buy diamonds where theprice is low [Europe] and ship them towhere the price is high * * * therebyeliminating the price difference [between]the two parts of the world.’’ Tr. at2157:16–19 (Elzinga).

But, according to Elzinga, arbitrage isnot a factor that can ‘‘stitch’’ the UnitedStates high function FMS and HRM mar-kets to the same markets in other parts ofthe world. Tr. at 2205:21 (Elzinga). Arbi-trage does not exist in the high functionFMS and HRM markets for two reasons,he testified. First, the products that con-sumers buy from Oracle, PeopleSoft orSAP are licensed products, accordingly,the consumers ‘‘do not have the legal au-thority’’ to resell the software to otherconsumers. Tr. at 2158:6 (Elzinga). Sec-ond, high function FMS and HRM is

tooled to ‘‘work * * * and meet the specif-ic configurations and capabilities [of onlyone consumer], it won’t work [on anotherconsumer’s computers].’’ Tr. at 2158:18–21 (Elzinga). Therefore, lack of the ‘‘arbi-trage factor’’ reinforced Elzinga’s proposi-tion that consumers cannot find substituteproducts outside of the United States, hetestified.

Prices in the United States are not af-fected by prices in other parts of the world.Elzinga posited that United States con-sumers of high function FMS and HRMcannot expect to be charged the sameprice that a European consumer is paying.Tr. at 2206:7–11 (Elzinga). ‘‘The UnitedStates is not affected by prices or outputof [high function FMS and HRM] outsidethe United States. And the flip side isalso true. [P]rices charged outside of theUnited States aren’t affected by pricescharged inside.’’ Tr. at 2206:10–12 (Elzin-ga).

Oracle’s Proposed Geographic Market

Oracle asks the court to reject the plain-tiffs’ proposed geographic market. Oracleargues that the geographic market in thiscase is ‘‘so clear[ly] [a global market] thatreasonable people ought not be debatingit.’’ Def Post Brief (Doc. # 365) at 22.Further, Oracle noted that this is not thefirst time the DOJ has tried (unsuccessful-ly) to claim a United States-only market inthe face of overwhelming evidence of aworldwide market. Def Post Brief (Doc.# 365) at 23 n. 19 (citing United States v.Eastman Kodak, 63 F.3d 95 (2d Cir.1995)).

Oracle assailed plaintiffs’ severance ofSAP into two distinct companies. ‘‘Theproposed United States-only market is away of * * * making SAP appear ‘smaller’than it really is and simultaneously makingOracle and PeopleSoft appear ‘bigger’ thanthey really are.’’ Def. Post Brief (Doc.# 365) at 23. While SAP America is re-sponsible for all sales of SAP software in


the United States and Canada, it sellssoftware that was manufactured in Germa-ny. SAP America has no North Americanmanufacturing sites. Def. Fact (Doc.# 357) ¶ 100 at 50. Further, all large dis-count rates offered to United States cus-tomers by SAP America must be approvedby SAP AG. Tr. at 2836:22–24 (Knowles).Accordingly, without SAP AG, SAP Amer-ica would (1) have nothing to sell and (2)not be able to offer competitive discounts.

Moreover, simply because SAP has alarger market share in Europe does notmean that the geographic market shouldbe limited to the United States. ‘‘Sharesare not determinative of how you definethe [geographic] market’’ Hausman testi-fied. SAP—all of SAP—must be includedhe stated.

Once SAP is seen as a single entity,defendant claims that there are four differ-ent ways of analyzing the geographic mar-ket in this case, all of which point to aworldwide market. Tr. at 3793:18–19(Hausman).

First, Hausman analyzed the geographicmarket under the ‘‘hypothetical monopo-list’’ test from the Guidelines. Tr. at3794:9–10 (Hausman). See Guidelines§ 1.21. Even assuming SAP America isdistinct from SAP AG, Hausman statedthat if a hypothetical monopolist in theUnited States imposed a SSNIP, SAP AGcould ‘‘of course hire plenty of salespeople* * * and come in and compete.’’ Tr. at3795:2–6 (Hausman). ‘‘[SAP AG’s] prod-uct would do just fine in the UnitedStates.’’ Id. Accordingly, ‘‘if [the court]looks at this market from a Merger Guide-lines approach, you need to look at this ona worldwide basis.’’ Tr. at 3795:1–12(Hausman).

Second, Hausman analyzed the geo-graphic market under the plaintiffs’ de-scription of the ‘‘high function needs’’ ofthe customers who buy high function soft-ware. Tr. at 3795:24–25 (Hausman).

Hausman described the DOJ’s productdefinition as ‘‘multi, multi, multi,’’ referringto the functionality that the DOJ claimshigh function software possesses. Tr. at3796:1–2 (Hausman). The software mustbe able to handle multiple currencies, frommultiple jurisdictions, while understandingmultiple languages. Different currenciesand different languages are clearly ‘‘inter-national or worldwide features,’’ and there-fore ‘‘bring a worldwide aspect’’ to theanalysis. Tr. at 3798:7–8, 18–20 (Haus-man).

Third, Hausman employed the E–H testthat was rejected by its own creator. Tr.at 3800–3804 (Hausman). Hausman statedthat this is a point that both he and Elzin-ga agree upon: the E–H test would onlybe satisfied if the geographic market weredefined worldwide. Tr. at 3801:7–11(Hausman). Hausman stated that Elzin-ga’s rejection of the E–H test was ‘‘inap-propriate’’ for two reasons. First, thereare several markets, other than the highfunction ERP market, where the clientbuys a ‘‘relationship’’ with the vendor (e.g.,the purchase of a mainframe computer orserver). But, it has never been arguedthat the computer market is not a world-wide market. Tr. at 3802:1–15 (Hausman).‘‘We see [this kind of relationship] in allsorts of high-technology markets. Yet,people agree that those are all world mar-kets.’’ Tr. at 3802:13–15 (Hausman). Sec-ond, the E–H test has ‘‘often’’ been appliedto several cases involving services basedupon a relationship with customers, suchas hospital merger cases. Tr. at 3803:12–20 (Hausman). Accordingly, the E–H testis appropriate for this type of relationship-oriented scenario as well, and all agreethat the E–H test mandates a worldwidemarket.

Finally, Hausman opined that there isempirical evidence showing that prices inEurope constrain prices in the United


States, and vice versa. Tr. at 3805 (Haus-man). Hausman studied the PeopleSoftDAFs submitted to the DOJ. He foundthat the average discount rates for People-Soft in the United States was 45.2 percent.Tr. at 3805:19 (Hausman). In Europe, theaverage discount was 45.1 percent. Tr. at3805:20 (Hausman). Hausman stated thatthese discount rates are ‘‘virtually identi-cal.’’ Tr. at 3805:22. If the competitiveconditions in Europe and the UnitedStates were wholly independent of eachother, one would expect to see completelydifferent discounts in both regions. Butthese facts demonstrate, in Hausman’sview, that the market needs to be analyzedon a global scale. Tr. at 3806 (Hausman).

Accordingly, Oracle urged the court tolook at concentration figures based upon aglobal market of all FMS and HRM soft-ware. Def. Fact (Doc. # 357) at 56. Us-ing these product and geographic marketdefinitions, Oracle offered the followingglobal FMS market shares: SAP, 19.2 per-cent; Oracle, 16.8 percent; and People-Soft, 12 percent. Ex. P0825 at 21. Amerged Oracle/PeopleSoft would, in Ora-cle’s view, possess a 28.8 percent marketshare in the FMS market.

For global HRM, Oracle offers the fol-lowing market shares: SAP, 11.9 percent;PeopleSoft, 11.3 percent; and Oracle, 3.04percent. Ex. D5815 at 9. A merged Ora-cle/PeopleSoft would possess only a 14.3percent market share in the HRM market.

Findings of Fact: Geographic Market

[14] The court finds that the relevantgeographic market (‘‘the area of effectivecompetition’’) in this case is a worldwidemarket. Tampa Electric Co. v. NashvilleCoal Co., 365 U.S. 320, 327–28, 81 S.Ct.623, 5 L.Ed.2d 580 (1961).

At the outset, the court must addressthe plaintiffs’ attempt to sever SAP intotwo companies—SAP America and SAPAG. The court finds this argument whollyunpersuasive. SAP America, while critical

to SAP’s success in North America, is notan independent company. This fact wasclearly shown by the testimony of Knowleswho stated that any large discount (usuallyabove 70 percent) that SAP America of-fers, clearly in the face of competition,must get that discount approved by SAPAG in Germany. Further, while thesource of the code is not determinative ofthis severance inquiry, it is important tonote that all of SAP America’s software ismanufactured and shipped from SAP AG.So without SAP AG, SAP America wouldhave nothing to sell, and even if it did haveits own manufacturing, SAP Americawould still have to get competitive discountrates approved by SAP AG. To view thesetwo dependent branches of SAP as sepa-rate entities would be asking the court toignore the reality of how the industrypresently operates. Accordingly, the courtfinds that SAP must be viewed as onesingle entity.

Next the court must decide the geo-graphic boundaries within which the mar-ket participants effectively compete. Thiscourt (per Judge Chesney) has used theE–H test in defining the relevant geo-graphic market for merger analysis. SeeState v. Sutter Health System, 84F.Supp.2d 1057, 1069 (N.D.Cal.2000) (‘‘Theanalytical process [of defining the geo-graphic market] generally begins with anapplication of the Elzinga–Hogarty test* * *.’’). Furthermore, the results of em-ploying the E–H test are undisputed. SeeTr. at 2155:9–10 (Elzinga) (admitting thatthe E–H test would dictate the court toview the market as a global market). El-zinga’s basis for rejecting the E–H test isunpersuasive. The court, while agreeingthat ‘‘relationships’’ are important in ERPsales, does not find that such relationshipsrender the E–H test inapplicable. First,the court can think of a number of salestransactions that involve marketing andnegotiation as well as installation and


maintenance ‘‘relationships’’ between sellerand vendor: computer sales, copier sales,motor vehicles to name a few. But toargue that these markets, all involving ma-jor foreign vendors, are limited to theUnited States would be untenable.

Second, the E–H test has been usedwhen important vendor-customer relation-ships are involved. A clear example isSutter Health, where Judge Chesney usedthe E–H test in a hospital merger case todetermine whether patients seeking medi-cal treatment outside of the Bay area re-quired a geographic market expansion.No one can argue that medical treatmentis not a ‘‘relationship’’ between the patientand the doctor. But the E–H test con-trolled the analysis, not the location of the‘‘relationship’’ between the doctor and thepatient. Further, Judge Chesney is notalone in applying the E–H test to hospitalmerger cases. See United States v. MercyHealth Services, 902 F.Supp. 968, 980(N.D.Iowa 1995).

Finally, the court cannot allow this ‘‘re-lationship’’ factor to solely dictate the geo-graphic boundaries for this case, as thecourt has already found that ‘‘non-relation-ship’’ solutions (i.e., outsourcing) cannot beexcluded from the product market.

Accordingly, the court holds that the E–H test is an appropriate method of deter-mining the ‘‘area of effective competition’’between vendors in this relevant market.Tampa Electric, 365 U.S. at 327, 81 S.Ct.623. Elzinga, creator of the test, admittedthat applying the E–H test would mandatea global market. The court thereforefinds that the geographic market in thiscase is global.

Findings of Fact: Market Sharesand Concentration

In addition to failing to meet their bur-den of proving a distinct product market,plaintiffs have failed to prove that the rele-vant product market in this case is geo-graphically bound to the United States.

Accordingly, the market share and concen-tration statistics presented by Elzinga arewholly inapplicable to the court’s analysis.The court is left with a new product mar-ket definition which includes, at least: (1)ERP sold by Oracle, SAP, PeopleSoft,Lawson, AMS and Microsoft; (2) outsourc-ing solutions; and (3) best of breed solu-tions. Further, this product market mustbe analyzed as a global one.

Not surprisingly, plaintiffs did not offerany market share data other than those ofElzinga. Oracle, while successfully pick-ing apart plaintiffs’ market definition didnot provide a definitive alternative of itsown. The only statistical data Oracle of-fered showed the 2002 global HRM sharesof Oracle, PeopleSoft and SAP, but did notinclude HRM data on AMS or Microsoft’sshare since the BearingPoint alliance.Moreover, Oracle offered even less in theway of FMS shares or concentration.

But it is plaintiffs, not defendant, whocarry the burden of proving market sharesand concentration in order to invoke thepresumptions of the case law or to sustaina showing in accordance with the Guide-lines. The court cannot furnish its ownstatistics.

Without the benefit of presumptions, theburden remains upon plaintiffs to comeforward with evidence of actual anticom-petitive effects.

Anticompetitive Effects

Plaintiffs’ Evidence of CoordinatedEffects

[15] Plaintiffs presented no evidence attrial on coordinated effects. This was awise decision, as proving the probability ofsuch collusion would definitely be an uphillbattle for two reasons. First, the productsin the relevant market are not homoge-neous. Plaintiffs themselves even argueagainst homogeneity. Pls. Post Brief(Doc. # 366) at 30 (stating that the prod-


ucts in the high function HRM and FMSmarket are ‘‘highly differentiated’’). Sec-ond, there is no price transparency in thismarket. Prices and discount rates forsoftware are known only to the vendor andthe customer, both of whom take greatpains to keep such information confiden-tial. Without homogeneity or transparen-cy, the market conditions are not condu-cive to coordinated effects, either tacit orexpress. Plaintiffs recognized this unlike-lihood. Id. at 38 (‘‘The fact that highfunction software is a differentiated prod-uct and that pricing is not transparentmake price coordination between Oracleand SAP unlikely.’’).

But in plaintiffs’ post-trial brief theyunexpectedly included an entire section ar-guing that a post-merger Oracle and SAPcould tacitly collude in allocating custom-ers or markets. Id. at 38–40. Plaintiffsargue that ‘‘Oracle is strong in the hightechnology and telecommunications’’ areawhile ‘‘SAP dominates the oil and gas in-dustry.’’ Id. at 39. Accordingly, Oracleand SAP could reach a tacit understandingbased upon ‘‘mutual trust and forbear-ance’’ and stop competing against eachother in those relevant areas. Pls. PostBrief at 38 (quoting Hospital Corp. of Am.,807 F.2d at 1391). But the court hassearched in vain for any testimony or ex-hibits regarding tacit territorial or marketdivisions by Oracle and SAP. With no evi-dence in the record regarding such a spec-ulative coordinated effects argument, thecourt finds this new theory to be withoutmerit.Plaintiffs’ Evidence of Unilateral Effects

Plaintiffs rest their theory of anticom-petitive effects on an attempt to prove thatOracle and PeopleSoft are in a ‘‘localized’’competition sphere (a ‘‘node’’) within thehigh function FMS and HRM market.This sphere does not include SAP or anyother vendors, and a merger of Oracle andPeopleSoft would, therefore, adversely af-

fect competition in this localized market.Pls. Post Brief (Doc. # 366) at 31–36; Tr.at 2448–2450 (McAfee). Plaintiffs also of-fered evidence to show that SAP could notreposition itself to replace the localizedcompetition that would allegedly be lost ifOracle and PeopleSoft merge. Pls. PostBrief (Doc. # 366) at 32–33.

In attempting to prove localized compe-tition between Oracle and PeopleSoft,plaintiffs relied on virtually the same kindof evidence used to prove the product mar-ket, including internal corporate docu-ments, SAP executive testimony, customerand consultant firm testimony and experttestimony.

Internal documents. Plaintiffs relyupon several quarterly ‘‘win/loss analysis’’documents that were compiled by Oracleduring 2003 to show that Oracle and Peo-pleSoft are each other’s ‘‘closest competi-tors.’’ Pls. Post Brief (Doc. # 366) at 31.In Quarter 1 of 2003, plaintiffs offeredevidence that Oracle lost to PeopleSoft 37percent of the time when the two were incompetition, while Oracle lost to SAP only15 percent of the time the two competed.Ex. P2090. Plaintiffs then offered evi-dence from Quarter 3 in which Oracle ‘‘ex-plicitly states’’ that ‘‘PeopleSoft is ourNumber # 1 competitor’’ and ‘‘SAP is ourNumber # 2 competitor.’’ Ex. P2093.

But what plaintiffs failed to mention re-garding the Quarter 3 findings is that Ora-cle lost to PeopleSoft 54 percent of thetime, while they lost to SAP 53 percent ofthe time. Accordingly, what separates the‘‘# 1 competitor’’ and ‘‘# 2 competitor’’ ofOracle is merely one percent. Ex. P2093.Moreover, these roughly equal loss ratioscontinued into Quarter 4 when Oracle lostto PeopleSoft 59 percent of the time, whilelosing to SAP 50 percent of the time. Ex.P2095. Accordingly, the court can drawno conclusions from the conflicting datawithin the win/loss reports upon which


plaintiffs focus. In fact, these documentsarguably negate a showing of localizationbetween Oracle and PeopleSoft more thanthey support such a finding.

SAP executive testimony. Plaintiffs at-tempt to localize PeopleSoft and Oracle byshowing that many customers have a nega-tive ‘‘perception’’ of SAP and that SAP isat a ‘‘substantial disadvantage’’ when itcomes to competing for customers in theUnited States (the geographic market thatthe court has already rejected). Pls. PostBrief (Doc. # 366) at 31–32. In provingthese negative perceptions, plaintiffs point-ed to the testimony of SAP America’sKnowles. At trial, Knowles agreed thatSAP has had to deal with ‘‘perceptions’’that SAP is ‘‘too costly and difficult toimplement.’’ Tr. at 2950:8–12 (Knowles).Further, plaintiffs cited evidence from con-sulting firms and Knowles stating thatSAP has had ‘‘trouble’’ breaking into cer-tain verticals in the United States. SeeEx. P3037 (Knowles dep 5/3/04) at Tr.67:21–68:7 (difficulty breaking into servicessector); Tr. at 1698:1–8 (Bass) (difficultyin entering banking industry).

In deciding the merits of this argument,the court is again perplexed by the incon-sistency within plaintiffs’ own evidence.In trying to prove Oracle and PeopleSoftare in localized competition, plaintiffs triedto downplay SAP’s presence in the UnitedStates and characterize SAP has being‘‘disadvantaged’’ and unable to enter sever-al markets. But plaintiffs’ own evidenceon market shares negates such a finding.Even assuming the relevant geographicmarket in this case was the United States,Elzinga’s calculations of market shares inso-called high function FMS has SAPranked highest (above Oracle and People-Soft) with a 39 percent market share. El-zinga demo # 10. Moreover, in the HRMhigh function market, plaintiffs’ expertranked SAP second with a 29 percent mar-ket share (beating Oracle). Elzinga demo

# 11. SAP is not a ‘‘disadvantaged’’ and‘‘troubled’’ competitor in the UnitedStates. If it were, SAP should not bebeating Oracle in both markets and beat-ing PeopleSoft in the FMS market. Ac-cordingly, the court cannot credit plain-tiffs’ argument that SAP is suffering fromnegative customer perceptions or is disad-vantaged in competing against Oracle andPeopleSoft.

Customer and consulting firm testimo-ny. In furtherance of this localization the-ory, plaintiffs argued that customer testi-mony shows that ‘‘Oracle and PeopleSoftpresent better alternatives in the UnitedStates than SAP.’’ Pls. Post Brief (Doc.# 366) at 32. Plaintiffs support this asser-tion by citing the testimony of five custom-ers who eliminated SAP from the finalround of negotiations and instead chose todeal with Oracle and PeopleSoft. Id. (cit-ing testimony of North Dakota, NeimanMarcus, Greyhound, AIMCO and Cox).

The court finds this evidence unpersua-sive for two reasons. First, the court can-not take the self-interested testimony offive companies which chose to eliminateSAP from consideration, and from thatsample draw the general conclusion thatSAP does not present a competitive alter-native to Oracle and PeopleSoft. Drawinggeneralized conclusions about an extreme-ly heterogeneous customer market basedupon testimony from a small sample is notonly unreliable, it is nearly impossible.See Sungard Data Sys., 172 F Supp 2d at182–83. Second, the most persuasive testi-mony from customers is not what they sayin court, but what they do in the market.And as Elzinga’s statistics showed, cus-tomers are buying SAP FMS more thanOracle and PeopleSoft FMS. Elzinga demo# 10. Customers are buying SAP HRMmore than that of Oracle. Elzinga demo# 11.


Plaintiffs rely upon two of the Big Fiveconsulting firms’ testimony stating ‘‘theybelieve SAP is often the third choice ofmany U.S. customers.’’ Pls Post Brief(Doc. # 366) at 32. According to Bearing-Point’s Keating, SAP has long been theleast flexible of the three vendors in theway it has sold its HRM and FMS soft-ware. Tr. at 901:6–20, 946:18–20 (Keat-ing). Also, Accenture’s Bass testified thatSAP was ‘‘less likely to discount than Ora-cle and PeopleSoft.’’ Pls Post Brief (Doc.# 366) at 32; P3198 (Bass Dep) at Tr.132:08–133:07. But the plaintiffs’ own evi-dence discounts this argument. While itmay be true that SAP has been the leastflexible and least likely to discount, theevidence introduced by Elzinga shows thatcustomers apparently are not deterred bySAP’s inflexibility or higher pricing. Cus-tomers still buy SAP software over Oracleand PeopleSoft. See Elzinga demo.## 10–11.

Taken as a whole, the customer andconsulting firm testimony falls short ofproving that Oracle and PeopleSoft engagein competition to which SAP is simply nota party. Moreover, both PeopleSoft indus-try witnesses conceded there is no verticalin which SAP is not competitive with Ora-cle and PeopleSoft. Tr. at 388:1–11(Bergquist); 1957:10–21 (Wilmington).

Expert testimony. Finally, plaintiffs of-fered the testimony of Professor McAfeeto show that PeopleSoft and Oracle areengaged in localized competition to whichSAP is not a party. McAfee conductedthree independent analyses to reach hisconclusions. Pls Post Brief (Doc. # 366)at 34.

First, McAfee examined, in detail, twen-ty-five of Oracle’s DAFs in which Oraclesalespersons had listed PeopleSoft as theirjustification for seeking a higher discount.Second, McAfee, using charts of discounttrends provided by Oracle, ran a regres-sion analysis to assess the effect of People-

Soft’s presence on Oracle’s discount levels.Third, using the market statistics calculat-ed by Elzinga, McAfee conducted a ‘‘merg-er simulation’’ to assess the theoreticaleffects of an Oracle/PeopleSoft merger.Tr. at 2447–2449 (McAfee). Based uponthese three independent studies, McAfeeconcluded that in many instances People-Soft and Oracle are each other’s closestcompetitor and a merger between the twowould cause significant anticompetitive ef-fects. Tr. at 2466:8–13, 2449:22–24 (McAf-ee).

Twenty-five case studies. At trial,McAfee showed the court several DAFs inwhich the presence of PeopleSoft had jus-tified an Oracle salesperson seeking asteep discount. McAfee then picked outexplicit language from the justification col-umn to prove that when Oracle and Peo-pleSoft compete, they do so vigorously.For example, when seeking a discount onthe Hallmark account, a salesperson’s jus-tification for a discount was an ‘‘EX-TREMELY competitive situation against’’PeopleSoft. Because of this competition, a‘‘higher discount was warranted.’’ Tr. at2464:15–21 (McAfee). Likewise, in tryingto win the Greyhound account Oraclewanted to cause a ‘‘third straight loss’’ forPeopleSoft and ‘‘only aggressive propos-als’’ would win Greyhound over. Tr. at2466:14–20 (McAfee).

These two examples are representativeof the many that McAfee showed thecourt—clear examples of how vigorouslyPeopleSoft and Oracle compete when theygo ‘‘head to head’’ against each other, heasserted. McAfee concluded that suchhead to head competition between Oracleand PeopleSoft would be lost if this merg-er were consummated. Tr. at 2488:13–25(McAfee).

Regressions. Next, in trying to showlocalized competition, McAfee used a re-gression technique to calculate what effect,


if any, the presence of PeopleSoft or an-other competitor has on the discounts of-fered by Oracle. Tr. at 2495:22–25 (McAf-ee). McAfee ran two regression analyses.In the first, McAfee was privy to salesrepresentative surveys identifying the dis-count percentages given to Oracle custom-ers that had purchased the E–BusinessSuite. The surveys also identified thecompetitor that Oracle had beaten to getthe account. Tr. at 2497:10–14 (McAfee).McAfee narrowed the sample to all salesthat were over $500,000, in order to equatethe sample with Elzinga’s market defini-tion. McAfee used these variables (com-petitor, net revenue, discount percentage)and ran the regressions. Tr. at 2498:3–20(McAfee). The data led McAfee to con-clude that ‘‘PeopleSoft has a .097 (9.7 per-cent) effect’’ on the discount Oracle offers.Tr. at 2499:22–25 (McAfee). In otherwords, when Oracle competes against Peo-pleSoft for the sale of Oracle’s E–BusinessSuite, the consumer obtains a 9.7 percentgreater discount than when Oracle com-petes against no one in selling the suite.

Wanting to look at more than just thesale of the E–Business suite, McAfee thenanalyzed all of the DAFs that Elzinga hadused in defining the product market andmatched those with the data from the salesrepresentative forms to create a largersample with more variables. Tr. at2504:22–25 (McAfee). The DAFs listedthe percentage requested along with thecompetitor justifying such a discount.Once McAfee ran this second regression,he concluded that PeopleSoft had a .136effect on Oracle’s discount rates (i.e., 13.6percent greater discount). Tr. at 2507:6–11 (McAfee). Accordingly, McAfee con-cluded that when PeopleSoft is competingagainst Oracle, Oracle’s discounts are 9 to14 percentage points greater. Tr. at 2508(McAfee).

Based upon these DAF studies and re-gression analyses, it is safe for the court to

conclude that Oracle and PeopleSoft docompete frequently for ERP customersand when they do compete, that competi-tion can be vigorous. But these two con-tentions are not disputed by anyone in thecase. Oracle concedes that PeopleSoft is afrequent rival. Def Post Brief (Doc.# 365) at 34. The court fails to under-stand what this undisputed fact is sup-posed to show about whether Oracle andPeopleSoft are competing head to head ina product space in which SAP is not aparty. McAfee himself stated that fromthese twenty-five DAFs, he drew the‘‘broad conclusion that in many instancesPeopleSoft and Oracle are each others’closest competitors.’’ Tr. at 2466:10–12(McAfee). But these DAFs tell the courtnothing about how often SAP competesagainst PeopleSoft or Oracle (a key factualissue if trying to exclude SAP) or whetherthat competition is equally fierce. Whatwould have been more helpful to the courtwould have been the DAFs of PeopleSoftand SAP as well. Defendants introducedseveral SAP DAFs during trial, one show-ing a very aggressive competition againstOracle, so it is clear that such forms exist.Ex D5649R A more complete DAF recordwould perhaps have evidenced localizedcompetition between Oracle and People-Soft. But plaintiffs did not provide suchDAFs to McAfee, nor is it clear whetherthey even sought to obtain such documentsduring discovery.

Simply because Oracle and PeopleSoftoften meet on the battlefield and fightaggressively does not lead to the conclu-sion that they do so in the absence of SAP.

Merger simulation. Finally, McAfeeconducted a merger simulation analysis.There are several merger simulation mod-els that can be used depending on thecharacteristics of the industry. Tr. at2511:12–19 (McAfee). McAfee chose the‘‘English auction’’ model (also called the


oral ascending auction) because one of thefeatures of this model is its allowance ofmultiple bidders and multiple rounds ofbidding. Tr. at 2511:19–22 (McAfee). Thesimulation works by putting in necessaryvariables and assumptions, such as marketshares and percentage of wins in head tohead competition. Once these variableswere accounted for, McAfee still had to seta variable for ‘‘how competitive the market[was] pre-merger.’’ Tr. at 2526:17–22(McAfee). One way of creating such ameasurement is by estimating the ‘‘totalvalue of the product that accrues to thebuyer’’ (i.e., how ‘‘much of the value of thesoftware to the buyer actually accrues tothe buyer and how much accrues to thevendors in the form of price’’). Tr. at2517:1–4 (McAfee). McAfee ran the sim-ulation based upon five different ‘‘buyeraccrual’’ estimates: .5 (only 50 percent ac-crual) to .9 (90 percent accrual). McAfeeused the market shares calculated by El-zinga as his market shares variable. Onceall the data are compiled and the variablesaccounted for, the merger is simulated bymerging the shares of the two mergingfirms. Once this is done, data can becalculated showing how much the price ofthe relevant product is expected to in-crease. McAfee asserted that in the highfunction FMS market, after the Ora-cle/PeopleSoft merger, he expects price toincrease anywhere from 5 percent (.50 ac-crual variable) to 11 percent (.90 accrualvariable). In the high function HRM mar-ket, McAfee concluded that the pricewould increase by 13 percent (.50 accrualvariable) to 30 percent (.90 accrual varia-ble). Ex. P4024.

McAfee asserted that this merger sim-ulation, using Elzinga’s market share sta-tistics, shows that a merger between Ora-cle and PeopleSoft will lead to a unilateralprice increase in both markets.

But the court has already found thatElzinga’s market share statistics are not a

reliable indicator of Oracle, SAP and Peo-pleSoft’s positions in the ERP market.Accordingly, because this merger sim-ulation is based upon these unreliable data,the court concludes that the simulationresults are likewise unreliable.

Oracle’s Competitive Effects Rebuttal

Oracle takes issue with all of the plain-tiffs’ evidence regarding the likelihood ofanticompetitive unilateral effects.

First, Oracle claims that the presentcase is not the type of case for which thedoctrine of unilateral effects was created.Oracle offered Campbell’s expert testimo-ny that a fundamental assumption of theunilateral effects theory is not present inthis case. Tr. at 2721:3–5 (Campbell).Campbell testified that the ‘‘unilateral ef-fects doctrine is posited on the notion of alocalized market powered by a seller and agroup of purchasers located in productspace or geographic space around thatparticular seller.’’ Tr. at 2721:5–9 (Camp-bell). This ‘‘product space’’ is defined bycharacteristics of the product or productswithin the space. Campbell offered ahomey example of product space usingbreakfast cereal. Tr. at 2721:15–18(Campbell). A number of customers havecharacteristic preferences for their break-fast cereal that could create a productspace within the entire breakfast cerealmarket. For example, some customersprefer cereal to be crunchy, sugar-free andhigh in fiber. These characteristics of theproduct will narrow the entire marketdown to a ‘‘space’’ in which only crunchy,sugarless, high fiber cereals occupy thespace and only those companies that pro-duce such cereal are competitors. Camp-bell called this space a ‘‘node,’’ with thebuyers being centered around this node.The unilateral effects theory is concernedabout there being only one vendor operat-ing inside the node, thereby being able toincrease the price unilaterally. Tr. at


2721:19–23 (Campbell). Plaintiffs attempt-ed to carve out a ‘‘node’’ for high functionFMS and HRM software in the UnitedStates in which only Oracle and PeopleSoftcompete. Accordingly, if a merger takesplace, there will be only one vendor in thisnode with the ability unilaterally to reduceoutput and raise price within the node.

Campbell asserted that the unilateraleffects theory is predicated on the funda-mental assumption that the consumers inthe node have no ‘‘buyer power.’’ He tes-tified that the theory assumes that custom-ers are unsophisticated buyers who willnot be able to rebuff a price increase. Tr.at 2721:23 (Campbell). This fundamentalassumption does not hold in the case of theproducts in suit. Campbell asserted thatthe buyers of high function FMS andHRM are extremely sophisticated andknowledgeable and engage in extensiveand intensive one-on-one negotiations withvendors. These customers clearly have alot of power during these negotiations,Campbell claimed, and they are aware ofthis power. Tr. at 2722:1–4 (Campbell).Campbell gave examples of high functionconsumers such as Emerson Electric andDaimler whose representatives testifiedthat their companies have ‘‘leverage’’ and‘‘power over people they deal with,’’ anduse their ‘‘size’’ and ‘‘the size of the deal’’to gain better deals on software. Camp-bell demo # 25 (citing Tr. at 1287:22–1288:2 (Peters); Tr. at 1407:20–1408:1(Gorriz)). Campbell concluded that theunilateral effects theory is ‘‘dogma devel-oped for a totally different context’’ fromthe present case. Tr. at 2728:6–7 (Camp-bell).

Even assuming arguendo that a unilat-eral effects theory is appropriate for thiscase, Oracle attacks each piece of evidencethat plaintiffs put forward attempting toprove localization between Oracle and Peo-pleSoft.

Oracle objected to plaintiffs’ character-ization of SAP as a struggling firm with asubstantial disadvantage which prevents itfrom being in a localized space with Oracleand PeopleSoft. Def. Post Brief (Doc.# 365) at 33. Oracle claims that theseSAP ‘‘struggling’’ assertions are ‘‘not re-motely true’’ and are belied by the factthat SAP has over 22,000 professional ser-vice customers. Id. While Oracle admitsthat SAP does not ‘‘dominate’’ the UnitedStates in the manner that it may ‘‘domi-nate elsewhere,’’ non-domination does notequate with ‘‘struggling.’’ Id.

Finally, Oracle takes aim at McAfee’sexpert testimony on anticompetitive ef-fects. First, Oracle claims that McAfee’s‘‘case studies’’ based upon the OracleDAFs do nothing more than ‘‘show Oracleand PeopleSoft are frequent rivals.’’ Id. at34. This evidence reveals nothing aboutlocalization between Oracle and PeopleSoftin a product space in which SAP is notencompassed. McAfee offered no insightsregarding the characteristics of high func-tion FMS and HRM that create the al-leged product space between Oracle andPeopleSoft. Further, these case studiesare devoid of any information about wheth-er head to head competition between Ora-cle and SAP, or PeopleSoft and SAP, isequally vigorous.

With regards to McAfee’s regressionanalysis, Oracle argued the analysis wasflawed from the outset. The data uponwhich McAfee based his regression were‘‘not based on any set of data identifying* * * high function HRM and FMS soft-ware, but only on data involving broadersuites of EAS.’’ Id. at 36. Accordingly, itis impossible to know if these alleged in-creased discount rates were the product ofhigh function FMS and HRM, other ERPpillars or the bundling of all. Without thiscrucial information, the regression analysisshows nothing in regards to localization


between Oracle and PeopleSoft in a highfunction FMS and HRM product space.Id. at 37.

Furthermore, Oracle assails McAfee’smerger simulation as ‘‘simplistic’’ and‘‘spurious.’’ Id. Oracle cites two majorflaws in the merger simulation. First, the‘‘auction’’ model is wholly inappropriate forthe present market because (1) the cus-tomers in this market are extremely pow-erful at bargaining and (2) vendors of ERPdo not simply ‘‘bid’’ for business; ratherthese negotiations are extensive and pro-longed, with the purchaser having com-plete control over information disclosure.Id. Second, the ‘‘market shares—the onlyinput having any connection to real-worlddata—were those produced by Elzinga us-ing the plaintiffs’ market definition.’’ Id.at 38. Because Oracle wholly rejectsplaintiffs’ ‘‘gerrymandered’’ market defini-tion, market statistics based upon this defi-nition are equally flawed. Accordingly,the merger simulation’s prediction of priceincreases after the merger are inaccurateand unreliable, based as it is on an inap-propriate model using inaccurate data.

Finally, and perhaps most importantly,Oracle contends that plaintiffs have offeredno ‘‘econometric calculations in trying toprove localization.’’ Id. at 31. Oracle ar-gues that proving localization requires ‘‘ex-tensive econometric analysis,’’ such as di-version ratios, price-cost margins and thelike, of which plaintiffs have offered none.When Oracle cross-examined plaintiffs’ ex-pert witnesses, both admitted that they‘‘did not even attempt to calculate diver-sion ratios, or cross-elasticities, or any oth-er economically meaningful measurementof whether the products of Oracle andPeopleSoft are uniquely close substitutesfor each other.’’ Id. See Tr. at 2293:23–25(Elzinga); Tr. at 2599:3–8 (McAfee).

Findings of Fact: Unilateral Effects

[16] The court finds that the plaintiffshave wholly failed to prove the fundamen-

tal aspect of a unilateral effects case—theyhave failed to show a ‘‘node’’ or an area oflocalized competition between Oracle andPeopleSoft. In other words, plaintiffshave failed to prove that there are a signif-icant number of customers (the ‘‘node’’)who regard Oracle and PeopleSoft as theirfirst and second choices. If plaintiffs hadmade such a showing, then the court couldanalyze the potential for exercise of mo-nopoly power over this ‘‘node’’ by a post-merger Oracle or the ability of SAP orLawson to reposition itself within the nodein order to constrain such an exercise ofmonopoly power.

Plaintiffs’ attempt to show localized com-petition based upon customer and experttestimony was flawed and unreliable.Moreover, plaintiffs’ evidence was devoidof any thorough econometric analysis suchas diversion ratios showing recapture ef-fects. Both the Kraft Gen Foods andSwedish Match courts, the only othercourts explicitly to address unilateral ef-fects, based their rulings in part upon eco-nometric evidence submitted by the par-ties. Kraft Gen. Foods, 926 F.Supp. at 356(relying on econometric evidence of thecross-price elasticity of demand betweenPost cereal brands and Nabisco brands);Swedish Match, 131 F Supp 2d at 169(relying upon the diversion ratio betweentwo brands of loose leaf tobacco).

Plaintiffs claim they were unable topresent the court with such econometricdata because ‘‘this [the high function HRMand FMS market] is a market that’s shotthrough with price discrimination,’’ andtherefore such data would be unreliable.Tr. at 2291:15–16 (Elzinga). But the courtfinds plaintiffs’ price discrimination argu-ment unpersuasive. First, ‘‘this’’ marketwhich Elzinga claims is plagued by pricediscrimination, is the so-called high func-tion FMS and HRM market that the courthas already rejected as being the relevant


product market in which to examine theeffects of the proposed merger. Second,assuming that the high function FMS andHRM market were the relevant market,which it clearly is not, plaintiffs only evi-dence regarding price discrimination camefrom Elzinga’s analysis of the OracleDAFs. Elzinga stated that there was awide range of discounts offered by Oracleto these 222 customers. Tr. at 2222:13–19(Elzinga). Elzinga stated that because Or-acle charged different discounts to thesecustomers, Oracle must be able to deter-mine what price it can charge a customerbefore the customer eliminates Oracle as apotential vendor (i.e., Oracle price discrim-inates). And since Oracle price discrimi-nates, then SAP and PeopleSoft must pricediscriminate as well.

But Elzinga admitted he conducted noformal studies of price discrimination in‘‘this’’ market. Tr. at 2343:14–20 (Elzin-ga). Nor did he examine the discountsgiven by PeopleSoft or SAP to their re-spective customers. Tr. at 2351:10–14(Elzinga). Elzinga’s assertion that thismarket is ‘‘shot through’’ with price dis-crimination because ‘‘somehow’’ Oraclewas able to determine what level of dis-count it could offer to different customersuncannily resembles his argument thatthere is ‘‘something different’’ about Ora-cle, PeopleSoft and SAP. Again, the courtrefuses to sustain plaintiffs’ inarticulablecontentions.

In sum, the court finds that plaintiffshave failed to show an area of localizedcompetition between Oracle and People-Soft.

Oracle’s Efficiency Defense

Oracle offers an efficiency defense torebut plaintiffs’ claim of anticompetitiveeffects. Def. Post Brief (Doc. # 365) at39–40. Oracle claims that the merger willresult in two overall efficiencies: (1) signif-icant cost-savings for Oracle in many areasof business, and (2) an increase in Oracle’s

scale (i.e., customer base), thereby fuelingmore competition with SAP, Siebel andMicrosoft resulting in higher innovationand lower costs. Def. Fact (Doc. # 357)¶¶ 234–237 at 113, ¶¶ 247–251 at 118–21.

Oracle’s cost-savings evidence camefrom a spreadsheet originally compiled inMay 2003 when Oracle wanted to acquireJ. D. Edwards. The spreadsheet was re-vamped in June 2003 when Oracle soughtto acquire PeopleSoft. It was finalized inJuly 2003 when Oracle looked at acquiringboth. Tr. at 3469:5–12, 3470:19–20 (Catz);Ex. D7132. (Acquisition Efficiencies Anal-ysis) (AEA). The AEA lists, as of July2003, PeopleSoft’s total costs for the areasof sales and marketing (S & M), researchand development (R & D) and general andadministrative (G & A). Id. For 2003,PeopleSoft’s total cost of S & M was $769.3million, R & D was $466.9 million and G &A costs were $214 million. Id. The AEAprojects that one year after Oracle hasacquired PeopleSoft, the cost of S & M willdecrease to $34 million ($735.3 million insavings), R & D will decrease to $201.3million ($265.6 million in savings) and G &A will decrease to $37.4 million ($176.6million in savings). Id. Accordingly, Ora-cle argues that post merger, it will achievecost-savings of over $1 billion. Def Fact(Doc. # 357) ¶ 234 at 113. Moreover, thecost savings are annual. So Oracle wouldsave $1 billion in 2005, $1 billion in 2006,and so forth. Tr. at 3493:2–5 (Catz).

Catz further testified to the efficienciesthat would result if Oracle’s scale wereexpanded to include PeopleSoft’s custom-ers. Tr. at 3438–3439 (Catz). Catz statedthat one of the main reasons, aside fromcost savings, that led Oracle to make atender offer for PeopleSoft was the poten-tial acquisition of PeopleSoft’s ‘‘customerbase.’’ Tr. at 3438:20 (Catz). The scale ofa company is a source of annuity revenue,revenue which allows a company to invest


more in research and development of itsproducts. Id. By acquiring PeopleSoft,Oracle would capture the extra revenue ofPeopleSoft’s customer base as well as thepotential for revenue from sales of add-onproducts. Tr. at 3439:6–12 (Catz). Thisadditional revenue and customer basewould allow Oracle to expand its R & D,thereby fueling more innovation of Oraclesoftware. Specifically, Catz testified abouta new ‘‘superset product line’’ that wouldhave the ‘‘best features from Oracle’’ andthe ‘‘best features and modules from Peo-pleSoft.’’ Tr. at 3451:2–7 (Catz).

Further, the larger customer base andincreased innovation would allow Oracle tocompete with larger competitors, such asMicrosoft, and compete better in otherERP markets, such as SCM and CMS. Tr.at 3440:3–7 (Catz); Def Fact (Doc. # 357)¶ 251. Reduced costs, increased innova-tion and more competition are efficienciesOracle claims outweigh, and thus rebut,any showing of anticompetitive effectsplaintiffs have put forward.

Plaintiffs’ Efficiency Rebuttal

Plaintiffs rebutted the efficiency defenseby calling Professor Zmijewski, a profes-sor of business from the University ofChicago. Zmijweski was asked to verifythe arithmetic in the AEA spreadsheetthat Oracle claims explicate its large cost-saving efficiencies. To verify the spread-sheet, Zmijewski was required to ‘‘teaseout’’ all of the inputs (i.e., the pre-mergercosts and the post-merger costs of all de-partments) that had been plugged in byOracle, verify that those inputs were true(based in fact) and then recalculate thenumbers to verify that the final efficiencyamounts were the same as the amountsrepresented on the AEA. Tr. at 4509:16,4517–4518 (Zmijewski).

Zmijewski teased out the inputs success-fully then began looking at informationprovided by Oracle and the SEC for some‘‘factual foundation’’ for these inputs and

post-merger assumptions Oracle had usedin calculating the AEA. Tr. at 4520:5 (Zmi-jewski). But Zmijewski hit a ‘‘dead end’’every time he tried to find some factualbasis for any of the inputs in the spread-sheet. Id. A four month search throughthe documents left Zmijewski with ‘‘essen-tially none’’ of the information he neededto verify the AEA inputs. Tr. at 4520:11(Zmijewski). Zmijewski’s uneasinessabout his fruitless search was relievedwhen he found that there was no factualbasis for the inputs. Catz had ‘‘used herpersonal judgment’’ based upon consulta-tion with Larry Ellison and others in de-termining the inputs that went into theAEA. Tr. at 4520:14–23 (Zmijewski);3558:1–8 (Catz). Further, there were nodocuments that could explain how Catzand others had reached these personaljudgments on the inputs. Tr. at 3558:21(Catz). This led Zmijewski to concludethat the AEA is ‘‘not verifiable’’ and there-fore not reliable under the verificationstandards used by many professionals, in-cluding the SEC. Tr. at 4519:24, 4516:5–12(Zmijewski). Plaintiffs claim that cost-sav-ing efficiencies require defendant to ‘‘ ‘ex-plain the methods used to calculate’ ’’ thecost-saving numbers. Pls Post Brief (Doc.# 366) at 47 (quoting Staples, 970 F.Supp.at 1089). According to plaintiffs, Oraclehas provided no explanation of the meth-ods used to calculate the AEA other thanthe judgment of Catz and her colleagues.

Finally, plaintiffs urge the court to putno stock in Oracle’s innovation claims, asthey are unverified and not merger-specif-ic. Pls. Post Brief (Doc. # 366) at 49–50.When Catz was cross-examined about thesuperset product line, the innovative hy-brid of Oracle and PeopleSoft, she did nothave any documents discussing this pro-posed innovation, nor did she know anydetails about when the product would beavailable. Pls. Post Brief (Doc. # 366) at50; Tr. at 3533:8–16 (Catz). Plaintiffs


claim this ‘‘vague’’ assertion of a supersetproduct line is not a cognizable innovationclaim under case law or the Guidelines.Pls. Post Brief (Doc. # 365) at 49 (citingHeinz, 246 F.3d at 723 (requiring ‘‘reliableand significant evidence that the mergerwill permit innovation that otherwise couldnot be accomplished * * *.’’)).

Findings of Fact: Efficiencies

[17] In order for a claimed efficiencyto be ‘‘cognizable,’’ it must be ‘‘substanti-ate[d]’’ and ‘‘verfi[able].’’ Guidelines § 4.0.The court finds Oracle’s evidence on theclaimed cost-savings efficiency to be flawedand unverifiable. Catz and Ellison’s per-sonal estimations regarding the potentialcost-savings to Oracle are much too specu-lative to be afforded credibility. Oracle’sefficiency defense based upon future inno-vations (e.g., the superset product) was notverified by internal documents. Oraclepresented no evidence regarding the func-tionality or characteristics the innovativeproduct will contain, nor any evidence re-garding its date of availability.

Accordingly, both claimed efficienciesare much too vague and unreliable to rebuta showing of anticompetitive effects.

Conclusions Of Law

This court has jurisdiction over this ac-tion pursuant to 28 USC §§ 1331, 1337(a)and 1345 and Section 15 of the ClaytonAct, 15 USC § 25. Venue is proper in thisdistrict pursuant to 15 USC § 22 and 28USC § 1391(c).

In order to succeed on their claim, plain-tiffs must prove by a preponderance of theevidence (1) the relevant product and geo-graphic market, and within this market (2)the effect of Oracle’s acquisition of People-Soft may be substantially to diminish com-petition. See Penn–Olin, 378 U.S. at 171,84 S.Ct. 1710.

Plaintiffs alleged a product market limit-ed to HRM and FMS software licensed byOracle, PeopleSoft and SAP. Plaintiffs also

alleged a geographic market limited to theUnited States.

Plaintiffs have proven that the relevantproduct market does not include incum-bent systems or the integration layer. Butplaintiffs failed to prove that outsourcingsolutions, best of breed solutions and so-called mid-market vendors should be ex-cluded from the relevant product market.Furthermore, plaintiffs have failed to es-tablish that the area of effective competi-tion is limited to the United States. Ac-cordingly, plaintiffs have failed to meettheir burden of proving the relevant mar-ket for section 7 analysis.

Because plaintiffs have failed to meetthis predicative burden, plaintiffs are notentitled to a presumption of illegality un-der Philadelphia Nat Bank or the Guide-lines.

Plaintiffs have failed to prove the likeli-hood that a post-merger Oracle and SAPwould tacitly coordinate by allocating cus-tomers or markets. Accordingly, theplaintiffs have not met their burden ofestablishing anticompetitive coordinatedeffects.

Plaintiffs have failed to prove an area oflocalized competition between Oracle andPeopleSoft in which a post-merger Oraclecould profitably impose a SSNIP. Accord-ingly, plaintiffs have not met their burdenof establishing the likelihood of anticom-petitive unilateral effects.

Notwithstanding that plaintiffs havefailed to carry their burden to be entitledto relief, Oracle has not proved by a pre-ponderance of the evidence cognizable effi-ciencies sufficient to rebut any anticom-petitive effects of Oracle’s acquisition ofPeopleSoft.

Because plaintiffs have not shown by apreponderance of the evidence that themerger of Oracle and PeopleSoft is likelysubstantially to lessen competition in a


relevant product and geographic market inviolation of 15 USC § 7, the court directsthe entry of judgment against plaintiffsand in favor of defendant Oracle Corpora-tion.

This order is stayed 10 days to permitplaintiffs to apply for appellate remedies.

IT IS SO ORDERED.

,

In re AIR CRASH OVER THETAIWAN STRAITS ON

MAY 25, 2002.

No. CV 03–3635 MMM (RNBx).

United States District Court,C.D. California.

July 20, 2004.Background: Heirs of 124 people killed inthe crash of an airliner brought actionagainst airline and aircraft manufacturerfor, inter alia, wrongful death, negligence,and strict products liability. Defendantsmoved to dismiss the actions regarding 121of the victims on forum non conveniensgrounds.Holdings: The District Court, Morrow, J.,held that(1) Taiwan was an adequate forum for

lawsuits, and(2) in light of defendants’ proffered stipu-

lation of liability if lawsuits were triedin Taiwan, a majority of private andpublic interest factors favored dismiss-al.

Motion granted.

1. Federal Courts O45Standard to be applied to a motion to

dismiss on forum non conveniens groundsis whether defendants have made a clearshowing of facts which establish such op-pression and vexation of a defendant as to

be out of proportion to plaintiff’s conven-ience, which may be shown to be slight ornonexistent.

2. Federal Courts O45Courts treat forum non conveniens as

an exceptional tool to be employed spar-ingly, and should not perceive it as a doc-trine that compels plaintiffs to choose theoptimal forum for their claim.

3. Federal Courts O45To obtain dismissal on forum non con-

veniens grounds, a defendant must demon-strate that an adequate alternative forumexists, and that private and public interestfactors favor trial there.

4. Federal Courts O45In demonstrating that an adequate al-

ternative forum exists, as would warrantdismissal of a lawsuit on forum non conve-niens grounds, relevant private interestsfavoring trial in a particular forum include(1) relative ease of access to sources ofproof, (2) availability of compulsory pro-cess for unwilling witnesses, (3) compara-tive cost of obtaining willing witnesses, (4)possibility of a view of any affected prem-ises, (5) ability to enforce any judgmenteventually obtained, and (6) all other prac-tical problems that make trial of a caseeasy, expeditious and inexpensive.

5. Federal Courts O45In demonstrating that an adequate al-

ternative forum exists, as would warrantdismissal of a lawsuit on forum non conve-niens grounds, relevant public interest fac-tors favoring trial in a particular foruminclude (1) court congestion, (2) the unfair-ness of burdening citizens in an unrelatedforum with jury duty, (3) the interest inhaving localized controversies decided athome, (4) the interest in trying the case ina forum familiar with the applicable law,and (5) the interest in avoiding unneces-sary conflicts of laws.

1098 331 FEDERAL SUPPLEMENT, 2d SERIES1098 331 FEDERAL SUPPLEMENT, 2d SERIES allegation of other...

Documents

Transcript of 1098 331 FEDERAL SUPPLEMENT, 2d SERIES1098 331 FEDERAL SUPPLEMENT, 2d SERIES allegation of other...