CENTURY 21 REAL ESTATE CORP. v. LENDINGTREE, INC. 211 · CENTURY 21 REAL ESTATE CORP. v....

211CENTURY 21 REAL ESTATE CORP. v. LENDINGTREE, INC.Cite as 425 F.3d 211 (3rd Cir. 2005)

v. S.S. Le Moyne D’Iberville, 372 F.2d 123,124 (2d Cir.1967). See Chappelle, 84 F.3dat 654–55 (citing with approval a panel ofthe Ninth Circuit in Dannenberg v. Soft-ware Toolworks Inc., 16 F.3d 1073, 1076–77 (9th Cir.1994)).

[4] The School voluntarily dismissedwithout prejudice its First Cause of Actionpursuant to Fed.R.Civ.P. 41(a)(2), whichprovides for dismissal of an action ‘‘atplaintiff’s insistence TTT upon such termsand conditions as the court deems proper.’’Fed.R.Civ.P. 41(a)(2) (‘‘Unless otherwisespecified in the order, a dismissal under[Fed.R.Civ.P. 41(a)(2) ] is without preju-dice.’’). The district court’s denial of theSchool’s Rule 54(b) motion explicitly recog-nizes that the dismissal of the First Causeof Action was both voluntary and withoutprejudice. And at oral argument of themotion to dismiss in this Court, the Schoolexpressly declined to abandon the claimwith prejudice.

The School argues that the rule in Chap-pelle is prudential rather than jurisdiction-al. See, e.g., Great Rivers Coop. v. Farm-land Indus., 198 F.3d 685, 689 (8th Cir.1999) (holding that ‘‘the question whetherparties will be permitted to ‘manufacture’appeals’’ by voluntarily dismissing theirclaims without prejudice ‘‘is not jurisdic-tional’’). As a matter of prudence, theSchool urges that we should overlook thecontrivance that brings this appeal to us,because doing so would allow this appealto be heard in tandem with the nearlyidentical issues already presented to us inthe Greylock appeal. The School’s argu-ment finds no support in the law of thisCircuit and we are not persuaded by thereasoning of the circuit to which the Schoolcites, see Great Rivers Coop., 198 F.3d at689. It may be more efficient to hear theGreylock appeal and this appeal simulta-neously, as the School contends; thenagain, it might have been still more effi-

cient to have heard the School as amicus inGreylock’s appeal taken from a final order.In any event, we read Chappelle as ajurisdictional rule, interpreting as it doesthe jurisdictional limitations imposed by 28U.S.C. § 1291.

For the foregoing reasons, the Prov-ince’s motion is granted and this appeal(docket number 05–1803) is dismissed.The School’s motion to refer the Province’smotion to dismiss to the merits panel isdenied as moot.

,

CENTURY 21 REAL ESTATE CORPO-RATION; Coldwell Banker Real Es-tate Corporation; Era Franchise Sys-tems, Inc.

v.

LENDINGTREE, INC., Appellant.

No. 03–4700.

United States Court of Appeals,Third Circuit.

Argued Dec. 6, 2004.

Oct. 11, 2005.

Background: Real estate companiesbrought trademark infringement actionagainst operator of Internet real estatereferral service that used the companiesmarks on its website. The United StatesDistrict Court for the District of New Jer-sey, Joseph A. Greenway, Jr., J., grantedthe plaintiffs’ motion for a preliminary in-junction, and defendant appealed.

Holdings: The Court of Appeals, Rendell,Circuit Judge, held that:

(1) changes defendant made to its websitedid not make the case moot;

212 425 FEDERAL REPORTER, 3d SERIES

(2) in nominative fair use cases, the plain-tiff must first prove that confusion islikely due to the defendant’s use ofplaintiff’s mark, and the burden thenshifts to defendant to show that itsnominative use of plaintiff’s mark isnonetheless fair; and

(3) remand was required to determine ifdefendant’s use of the plaintiffs’ trade-marks was likely to cause consumerconfusion, and if so, whether such useof the marks amounted to nominal fairuse.

Reversed and remanded.

Fisher, Circuit Judge, filed opinion concur-ring in part and dissenting in part.

1. Trademarks O1523(3)‘‘Nominative fair use’’ is said to occur

when the alleged infringer uses the trade-mark holder’s product, even if the allegedinfringer’s ultimate goal is to describe hisown product; nominative fair use also oc-curs if the only practical way to refer tosomething is to use the trademarked term.

See publication Words and Phras-es for other judicial constructionsand definitions.

2. Trademarks O1523(2)‘‘Classic fair use’’ occurs where the

defendant uses the plaintiff’s mark to de-scribe the defendant’s own product.

See publication Words and Phras-es for other judicial constructionsand definitions.

3. Federal Courts O12.1Although a case may become moot if

the defendant can demonstrate that thereis no reasonable expectation that thewrong will be repeated, this burden is aheavy one.

4. Federal Courts O13Changes made by operator of Internet

real estate referral network to allegedly

infringing language on its website did notmoot trademark infringement actionbrought by real estate companies referredto in the website; modified language wasstill potentially infringing because it stillreferenced the companies’ marks, and find-ing of mootness would have permitted op-erator to return to its old ways after thethreat of a lawsuit had passed.

5. Trademarks O1421Likelihood of confusion is an essential

indicator of whether or not trademark in-fringement has occurred. Lanham Trade–Mark Act, §§ 32, 43(a), 15 U.S.C.A.§§ 1114(1), 1125(a).

6. Trademarks O1609, 1612Burden of proving likelihood of confu-

sion should remain with the plaintiff in atrademark infringement case, includingone where the defendant claims nominativefair use. Lanham Trade–Mark Act, §§ 32,43(a), 15 U.S.C.A. §§ 1114(1), 1125(a).

7. Trademarks O1609, 1612In nominative fair use cases, the plain-

tiff must first prove that confusion is likelydue to the defendant’s use of plaintiff’smark; once plaintiff has met its burden ofproving that confusion is likely, the burdenthen shifts to defendant to show that itsnominative use of plaintiff’s mark is none-theless fair. Lanham Trade–Mark Act,§§ 32, 43(a), 15 U.S.C.A. §§ 1114(1),1125(a).

8. Trademarks O1523(3)To demonstrate fairness, defendant

claiming nominative fair use of mark mustshow: (1) that the use of plaintiff’s mark isnecessary to describe both the plaintiff’sproduct or service and the defendant’sproduct or service; (2) that the defendantuses only so much of the plaintiff’s mark asis necessary to describe plaintiff’s product;and (3) that the defendant’s conduct orlanguage reflect the true and accurate re-


lationship between plaintiff and defen-dant’s products or services. LanhamTrade–Mark Act, §§ 32, 43(a), 15 U.S.C.A.§§ 1114(1), 1125(a).

9. Trademarks O1081

Traditional likelihood of confusion testassesses the following: (1) degree of simi-larity between owner’s mark and allegedinfringing mark; (2) strength of owner’smark;(3) price of goods and other factorsindicative of care and attention expected ofconsumers when making a purchase; (4)length of time defendant has used themark without evidence of actual confusion;(5) intent of defendant in adopting themark; (6) evidence of actual confusion; (7)whether the goods, though not competing,are marketed through the same channelsof trade and advertised through the samemedia;(8) extent to which the targets ofthe parties’ sales efforts are the same; (9)relationship of the goods in the minds ofconsumers because of similarity of func-tion; and (10) other facts suggesting thatthe consuming public might expect priorowner to manufacture a product in defen-dant’s market or that he is likely to expandinto that market. Lanham Trade–MarkAct, §§ 32, 43(a), 15 U.S.C.A. §§ 1114(1),1125(a).

10. Trademarks O1086, 1111, 1112,1523(3)

Factors most relevant in assessinglikely confusion in nominative fair use caseinclude: (1) the price of the goods andother factors indicative of the care andattention expected of consumers whenmaking a purchase; (2) the length of timethe defendant has used the mark withoutevidence of actual confusion; (3) the intentof the defendant in adopting the mark;and (4) the evidence of actual confusion.Lanham Trade–Mark Act, §§ 32, 43(a), 15U.S.C.A. §§ 1114(1), 1125(a).

11. Federal Courts O947

Remand was required in trademarkinfringement action to determine if use ofreal estate companies’ trademarks by op-erator of Internet real estate referral net-work was likely to cause consumer confu-sion, and if so, whether such use of themarks amounted to nominative fair use.Lanham Trade–Mark Act, §§ 32, 43(a), 15U.S.C.A. §§ 1114(1), 1125(a).

12. Trademarks O1111, 1523(3)

Relevant question in context of deter-mining defendant’s intent in adoptingplaintiff’s mark when determining likeli-hood of confusion in nominative fair usecase is not whether the defendant intendedto use the plaintiff’s mark, which it alwayshas in a fair use case, but whether it usedthe mark with the intent to confuse thepublic as to the relationship between thedefendant and the plaintiff; use of themark alone is not sufficiently probative ofsuch intent.

13. Trademarks O1113, 1523(3)

Disclaimers must be considered innominative fair use case in determiningwhether alleged trademark infringer accu-rately portrayed the relationship that ex-isted between plaintiff and defendant.Lanham Trade–Mark Act, §§ 32, 43(a), 15U.S.C.A. §§ 1114(1), 1125(a).

Bruce I. Goldstein, Saiber, Schlesinger,Satz & Goldstein, Newark, NJ, Jeffrey A.Conciatori [Argued], Quinn, Emanuel, Ur-quhart, Oliver & Hedges, New York, NY,for Appellant LendingTree, Inc.

Stephen W. Feingold [Argued], PitneyHardin, New York, NY, for Appellees Cen-tury 21 Real Estate Corporation; ColdwellBanker Real Estate Corporation; EraFranchise Systems, Inc.


Before RENDELL, FISHER, CircuitJudges, and YOHN,* District Judge.

FISHER, Circuit Judge, concurring inpart and dissenting in part.

OPINION OF THE COURT

RENDELL, Circuit Judge.

This case presents an opportunity for usto consider the contours of the traditionaltest for trademark infringement where thedefendant asserts the defense of ‘‘nomi-native fair use.’’ More specifically, wemust determine what role likelihood ofconfusion plays in a trademark infringe-ment case where the defendant claims thatits use was nominative and fair.

Appellees, Century 21, Coldwell Bankerand ERA (‘‘CCE’’) complain that AppellantLending Tree (‘‘LT’’), in the process ofmarketing its mortgage services, improp-erly referenced CCE’s trademarked ser-vices. LT contends that its use was nomi-native and fair, and permitted as a matterof law.

[1, 2] ‘‘Nominative’’ fair use is said tooccur ‘‘when the alleged infringer uses the[trademark holder’s] product, even if thealleged infringer’s ultimate goal is to de-scribe his own product. Nominative fairuse also occurs if the only practical way torefer to something is to use the trade-marked term.’’ KP Permanent Make–Up,Inc. v. Lasting Impression I, Inc., 328F.3d 1061, 1072 (9th Cir.2003) (rev’d. onother grounds) (quotations omitted). Bycontrast, ‘‘classic’’ fair use occurs wherethe defendant uses the plaintiff’s mark todescribe the defendant’s own product.New Kids on the Block v. News AmericaPub., Inc., 971 F.2d 302, 308 (9th Cir.1992).

The use of the term ‘‘Volkswagen’’ by acar mechanic in an ad describing the typesof cars he repairs has been held to consti-tute a nominative fair use. See id. at 307(citing Volkswagenwerk Aktiengesellschaftv. Church, 411 F.2d 350 (1969)). Clearly,the mechanic is referring to another’sproduct, but does so in order to describewhat he does. On the other hand, the useof the term ‘‘micro-colors,’’ a registeredtrademark of one make-up company, refer-ring to the pigments of the product of adifferent and competing make-up companythat it used in its own product, was classi-fied as a classic fair use. See KP Perma-nent Make–Up, Inc., 328 F.3d at 1072.There, the reference to the mark of anoth-er was made in describing its own productand its attributes.

Traditionally, we have looked to whethera trademark is likely to cause confusion inorder to determine whether a violation ofthe Lanham Act has occurred and, thus,whether the use should be enjoined andprohibited. However, it is unclear whatrole ‘‘likelihood of confusion’’ plays in theanalysis when ‘‘fair use’’ is asserted as adefense. Recently, the United States Su-preme Court provided guidance to thecourts regarding the test for classic fairuse in KP Permanent Make–Up, Inc. v.Lasting Impression I, Inc., 543 U.S. 111,125 S.Ct. 542, 160 L.Ed.2d 440 (2004).The issue before us is the extent to whichits reasoning applies to the nominative fairuse analysis as well.

I. Factual and Procedural Background

Appellees Century 21 and ERA haveeach been in business for over 30 years.Coldwell Banker has been in business foralmost 100 years. Each of these real es-tate companies oversees a system of fran-

* Hon. William H. Yohn, Jr., Judge of the Unit-ed States District Court for the Eastern Dis-

trict of Pennsylvania, sitting by designation.


chisees who operate by reference to thefranchisor’s trademark (e.g., Century 21Smith Realty). Franchise agreementspermit those brokerage companies to pro-vide realty services under trademarks heldby Cendant Corporation. Every franchi-see is granted a license to use its franchi-sor’s trademark only in connection with its‘‘d/b/a’’ name. (Thus, a franchisee couldonly refer to itself as Century 21 SmithRealty and not just Century 21). Thereare over 8,200 franchisees in CCE’s collec-tive systems in the U.S.

Appellant LendingTree describes itselfas a diversified consumer-oriented Inter-net business that helps consumers identifyand select qualified lenders, real estatebrokers, auto insurers, and other financialservice companies. It has a real estatereferral service that consumers can accessby visiting its website and inputting thelocation and characteristics of the housethey are seeking to buy or sell. LT thenselects and transmits information about upto four real estate companies participatingin LT’s referral network that service thatcommunity. If consumers ultimately usean LT referred broker, they receive re-wards, such as airline frequent flier milesand gift cards. LT also has an establishedmortgage referral program based on con-tractual relationships with participating fi-nancial institutions identified on its web-site.

At the time of this action, LT’s realestate referral network consisted of morethan 650 real estate broker member com-panies in the U.S., which collectively oper-ated more than 2,500 offices. More than9,000 real estate agents in those officeswere registered participants in LT’s net-work. At least 257 of the approximately650 real estate broker member companiesparticipating in the LT network operated aCentury 21, Coldwell Banker, or ERAfranchise.

LT’s real estate referral network wasformed in 1998 by HomeSpace, a companyfrom which LT acquired certain assets inAugust 2000. As early as 1999, HomeS-pace publicly advertised its real estate re-ferral network in printed materials as ‘‘in-cluding brokers representing the nation’sleading franchises, such as Coldwell Bank-er, Century 21 TTT and ERA.’’

The alleged infringement here is basedon the following uses of CCE’s marks:

(1) A Coldwell Banker ‘‘For Sale’’ signwith a woman, purporting to be a realestate agent, next to it, on which the blueand white Coldwell Banker logo was some-what obscured by the word ‘‘SOLD.’’ LT’sphone number was at the bottom. Thisscene was depicted at the bottom of LT’shomepage on its website.

(2) A statement by LT on its ‘‘Find aRealtor’’ homepage stating that LT will‘‘give you access to a national network ofbrokers representing the country’s leadingreal estate companies, including ColdwellBanker, ERA and Century 21.’’ Thesethree names headed a bullet-pointed list ofall such realtors to whom LT promisedaccess. The marks on those pages were inblock letter format.

(3) LT’s statement on its website’s HelpCenter that LT is ‘‘[r]epresented by largeindependent real estate companies andmembers of major franchises—ColdwellBanker, Century 21, Prudential, ERA, Re-MAX, GMAC (formerly Better Homes &Gardens), and Realty Executives.’’

(4) LT’s use of printed marketing mate-rials that stated that ‘‘LT is affiliated withmore than 700 certified brokers such asColdwell Banker, Century 21, Prudential,ERA and RE/MAX.’’

In January 2003, Kathryn Geib, in-housecounsel for CCE’s parent company respon-sible for the enforcement of its trade-marks, wrote to LT to demand that it stop


using CCE’s ‘‘marks’’ on its website in anymanner in the operation of its business.At that time, LT was using CCE’s logos onits website. After receipt of the letter, LTstopped using the logos (or any other ofCCE’s marks) on that webpage, but con-tinued to use CCE’s marks in block letterform on other webpages. In March 2003,Geib learned that LT was using a market-ing coupon containing the ‘‘affiliated lan-guage’’ described above. Geib sent a let-ter asking LT to stop such use. In May2003, CCE discovered that LT was usingCCE’s marks, but in block letters, on itswebpage. Geib again wrote, asking LT tostop such use.

Not satisfied with the response from LT,CCE commenced this action and appliedfor a preliminary injunction against LT’suse of its marks, claiming unfair competi-tion and trademark infringement in viola-tion of §§ 32 and 43(a) of the Lanham Act,15 U.S.C. §§ 1114(1) and 1125(a). At issuein the case before the District Court wasthe use of CCE’s marks in block form inbullet points on LT’s webpage, marketingmaterials using the ‘‘affiliated with’’ lan-guage described above, and the ColdwellBanker ‘‘For Sale’’ sign.

After oral argument before the DistrictCourt, LT voluntarily implemented certainmodifications to its website. It changedthe background color of the ‘‘For Sale’’sign from blue to red and removed thephone number at the bottom of the sign,moved the position of the bullet pointswith plaintiffs’ names on them from first tolast in its ‘‘Find a Realtor’’ list, and addeda disclaimer to its real estate homepagethat ‘‘LT is not sponsored by or affiliatedwith the parent franchisor companies ofany of the participating members of itsnetwork.’’ LT also added a note to a pop-up screen saying that ‘‘A particular fran-chisee in your geographic area may not beavailable to you through the LendingTree

network if that franchisee is not a partici-pating member of the network. In addi-tion, it is possible that the network maynot include any franchisee of a particularnational company in a given area.’’ (Textfrom www.LendingTree.com at JA689.)LT notified the District Court of thesechanges and represented that all of thelanguage allegedly giving rise to CCE’smotion had been permanently removedfrom LT’s website and would not be usedin the future.

The District Court issued its ruling withrespect to the language used by LT at thetime the complaint was filed, as well as themodified language. The District Court de-termined that LT’s use of Appellees’names was likely to cause consumer confu-sion, that the nominative use defense didnot shield LT in this instance, and thus,granted CCE’s motion for a preliminaryinjunction as to the use of CCE’s marksand the ‘‘affiliation’’ language, but not the‘‘For Sale’’ sign. The District Court’sanalysis did not separate out the originallanguage from the later modified language,as the District Court seemed to concludethat the revisions did not alleviate theproblem. LT now appeals this issuance ofthe preliminary injunction.

On appeal, we must first decide whetherthe case was moot when the District Courtissued its ruling, given the fact that LThad made several changes to the allegedlyinfringing language and the picture on itswebsite. If the case was not moot, wemust then determine the proper test toapply in the situation where a defendant isasserting a nominative fair use defense toclaims of unfair competition and trade-mark infringement under the Lanham Act.

II. Jurisdiction

The District Court had jurisdiction tohear this case pursuant to 28 U.S.C.§§ 1331 and 1338. We have jurisdiction


over this interlocutory appeal of a prelimi-nary injunction under 28 U.S.C.§ 1292(a)(1).

III. Mootness

[3, 4] Although a case may becomemoot ‘‘if the defendant can demonstratethat there is no reasonable expectationthat the wrong will be repeated, this bur-den is a heavy one TTT’’ and has not beenmet here. United States v. W.T. GrantCo., 345 U.S. 629, 633, 73 S.Ct. 894, 97L.Ed. 1303 (1953) (quotation omitted).The modified language that LT employedstill referenced CCE’s marks and, thus, ispotentially infringing and it cannot be saidthat LT’s voluntary discontinued use ofcertain language completely obviates theneed for injunctive relief. See LyonsPartnership v. Morris Costumes, Inc., 243F.3d 789, 800 (4th Cir.2001) (citation omit-ted). For instance, the use of CCE’smarks in block letter format could consti-tute trademark infringement in certain cir-cumstances. Whether these circumstancesare present in this case is still a questionthat needs to be resolved.

Furthermore, were we to hold that LT’svoluntary cessation of the alleged infring-ing activities rendered the case moot, thiswould potentially mean that LT would‘‘simply be free to return to [its] old waysafter the threat of a lawsuit had passed.’’Iron Arrow Honor Soc. v. Heckler, 464U.S. 67, 72, 104 S.Ct. 373, 78 L.Ed.2d 58(1983) (quotation omitted). Therefore, thecase is not moot and we will proceed toanalyze the substantive claims of trade-mark infringement and unfair competition.

IV. Fair Use

It must be recognized at the outset that‘‘fair use’’ presents a fact pattern differentfrom that of a normal infringement suit.The typical situation in a trademark caseinvolves the defendant’s having passed off

another’s mark as its own or having used asimilar name, confusing the public as toprecisely whose goods are being sold.Likelihood of confusion is the sole issue.But the fair use defense, by reason of thecircumstances giving rise to its applicabili-ty, alters the premise somewhat. The de-fendant is not purporting to be sellinggoods or services that the plaintiff hastrademarked, but, rather, is using plain-tiff’s mark in order to refer to defendant’sown goods or to the goods of the trade-mark owner in a way that might confusethe public as to the relationship betweenthe two. See Cairns v. Franklin Mint Co.,292 F.3d 1139, 1151 (9th Cir.2002). Ac-cordingly, the legal framework still in-volves a showing that A’s reference to B’smark will likely confuse the public, but theanalysis does not end there, for the usemay nonetheless be permissible if it is‘‘fair.’’

In KP Permanent Make–Up, Inc. v.Lasting Impression I, Inc., 543 U.S. 111,125 S.Ct. 542, 545–46, 160 L.Ed.2d 440(2004), the Supreme Court rejected thenotion that, in the context of classic fairuse, the party asserting the fair use de-fense to a claim of trademark infringementhad any burden to negate the likelihoodthat the practice complained of will con-fuse consumers about the origin of theservices or goods affected. Instead, plain-tiff has the exclusive burden to demon-strate likelihood of confusion, and thendefendant’s burden is only to show theaffirmative defense of fair use. The Su-preme Court stated, ‘‘[s]ince the burden ofproving likelihood of confusion rests withthe plaintiff, and the fair use defendanthas no free-standing need to show confu-sion unlikely TTT it follows that some possi-bility of consumer confusion must be com-patible with fair useTTTT’’ Id. at 550, 125S.Ct. 542. Thus, consumer confusion andfair use are not mutually exclusive. The


latter will in essence rebut or excuse theformer so that the use is permissible.

Before the Supreme Court spoke on theissue of classic fair use, the Court of Ap-peals for the Ninth Circuit had charted apath through a different fair use analy-sis—nominative fair use. In New Kids onthe Block, 971 F.2d at 308, the Court ofAppeals for the Ninth Circuit adopted itsown test governing the nominative fair useanalysis where the marks are used, as theyare here and in the case of the mechanic’sad described above that referenced Volks-wagen, to refer to the plaintiff trademarkowner’s product in order to help betterdescribe the defendant’s product or ser-vice. Thereafter, it refined the test andclarified its application. See, e.g., PlayboyEnterprises, Inc. v. Welles, 279 F.3d 796(9th Cir.2002) (holding that the use of theterms ‘‘Playboy,’’ ‘‘Playmate,’’ and ‘‘Play-mate of the Year 1981’’ on the websitemasthead and banner ads and in the meta-tags of the website of a former Playmateof the Year were nominative fair uses be-cause they served to identify the defendantand did not imply current sponsorship orendorsement); Cairns v. Franklin MintCo., 292 F.3d 1139 (9th Cir.2002) (holdingthat the sale of collectibles bearing thename and likeness of Princess Diana was anominative fair use).

Few other courts have spoken on theprecise issue of how nominative fair use issuccessfully invoked. Indeed, it seemsthat only the Second, Fifth, and Sixth Cir-cuits have referenced the nominative fairuse defense by name and even on theseoccasions have done so only to refer towhat district courts had done with theissue or to decline to adopt the NinthCircuit’s test as a whole. See Chambers v.Time Warner, Inc., 282 F.3d 147, 156 (2dCir.2002) (noting that the district courthad applied the standard for nominativefair use as articulated by the Ninth Cir-

cuit, but finding that the court had erredin its application); see also Pebble BeachCo. v. Tour 18 I Ltd., 155 F.3d 526, 546 n.13 (5th Cir.1998) (adopting the Ninth Cir-cuit’s nominative fair use test in part);Interactive Products Corp. v. a2z MobileOffice Solutions, Inc., 326 F.3d 687, 698 n.6 (6th Cir.2003) (footnoting why a districtcourt case involving the nominative fairuse defense was distinguishable from thecase before it); PACCAR Inc. v. TeleScanTechnologies, L.L.C., 319 F.3d 243, 256(6th Cir.2003) (declining to follow theNinth Circuit’s analysis (formulated priorto the Supreme Court’s holding in KPPermanent Make–Up ) and applying thelikelihood of confusion test instead).

Furthermore, this issue is one of firstimpression in our Court. See Basic Fun,Inc. v. X–Concepts, 157 F.Supp.2d 449, 456(E.D.Pa.2001) (stating that ‘‘the Third Cir-cuit does not recognize the ‘nominative’fair use defense, which is utilized as adefense solely in the Ninth Circuit’’). TheUnited States District Court for the Dis-trict of New Jersey did venture into thearea of nominative fair use defense in Liq-uid Glass Enterprises, Inc. v. Dr. Ingh.c.F. Porsche AG and Porsche CarsNorth America, Inc., 8 F.Supp.2d 398(D.N.J.1998). That case involved the useof Porsche’s trademarks by Liquid Glass, acorporation that sold car care products.Liquid Glass used Porsche’s trademarks inits advertisements for car polish—in par-ticular, an ad involving a provocativelydressed woman polishing a Porsche 911with the trademark ‘‘PORSCHE’’ plainlyvisible. The court initially analyzed thecase under the nominative fair use defensearticulated by the Ninth Circuit Court ofAppeals in New Kids On The Block, how-ever, it conflated the test with the plain-tiff’s burden of proving likelihood of confu-sion. See Liquid Glass, 8 F.Supp.2d at403 (finding that Liquid Glass’ fair usedefense was defeated, in part, because


Porsche would ‘‘likely prevail in demon-strating that the use of Porsche’s trade-marks TTT in Liquid Glass’s advertise-ments would likely confuse consumers asto Porsche’s connection with LiquidGlass’’). The Court reviewed each of thetraditional factors used to determine thelikelihood of confusion, though this analy-sis is not even a part of the New Kids OnThe Block test. This court’s difficulty inapplying the traditional test for likelihoodof confusion together with the nominativefair use defense highlights the need toclarify the proper analysis in this area ofthe law. Given the paucity of the case lawon this subject from every court other thanthe Court of Appeals for the Ninth Circuit,and that Court’s extensive study of thisissue, we will consider the opinions of thatCourt.

In New Kids On The Block, the Court ofAppeals for the Ninth Circuit consideredwhether two separate newspapers that hadconducted polls asking readers to vote fortheir favorite New Kid On the Block, andcharging money for each telephone votecast, had employed the ‘‘New Kids’’ trade-mark in such a way as to violate theLanham Act. The trademarked name wasvery valuable to the New Kids, as it hadbeen used to sell over 500 different prod-ucts, the profits of which partially inuredto the benefit of the New Kids. New Kidsalleged that the use of the name in con-ducting the polls implied that the NewKids were sponsoring the polls.

The Court began with a discussion ofthe history of trademark law, noting that:

[a] trademark is a limited propertyright in a particular word, phrase orsymbol. And although the English lan-guage is a language rich in imagery, weneed not belabor the point that somewords, phrases or symbols better conveytheir intended meanings than others TTT

Thus, trademark law recognizes a de-

fense where the mark is used only todescribe the goods or services of [a]party, or their geographic origin. 15U.S.C. § 1115(b)(4). The ‘fair-use’ de-fense, in essence, forbids a trademarkregistrant to appropriate a descriptiveterm for his exclusive use and so pre-vent others from accurately describing acharacteristic of their goods.

New Kids On The Block, 971 F.2d at 306(quoting Soweco, Inc. v. Shell Oil Co., 617F.2d 1178, 1185 (5th Cir.1980)).

The court reasoned that nominative fairuse cases—‘‘the class of cases where theuse of the trademark does not attempt tocapitalize on consumer confusion or to ap-propriate the cachet of one product for adifferent one’’—are not the typical fair usecases. It described the Volkswagen use asan example of this atypical type of fair use:

A good example of this is Volkswagen-werk Aktiengesellschaft v. Church, 411F.2d 350 (9th Cir.1969), where we heldthat Volkswagen could not prevent anautomobile repair shop from using itsmark. We recognized that in ‘‘advertis-ing [the repair of Volkswagens, it] wouldbe difficult, if not impossible, for[Church] to avoid altogether the use ofthe word ‘Volkswagen’ or its abbrevia-tion ‘VW,’ which are the normal termswhich, to the public at large, signifyappellant’s cars.’’ Id. at 352. Churchdid not suggest to customers that hewas part of the Volkswagen organizationor that his repair shop was sponsored orauthorized by VW; he merely used thewords ‘‘Volkswagen’’ and ‘‘VW’’ to con-vey information about the types of carshe repaired. Therefore, his use of theVolkswagen trademark was not an in-fringing use.

New Kids On The Block, 971 F.2d at 307.

The Court then reasoned that such asituation should be viewed as ‘‘involving anon-trademark use of a mark TTT’’:


Such nominative use of a mark—wherethe only word reasonably available todescribe a particular thing is pressedinto service—lies outside the stricturesof trademark law: Because it does notimplicate the source-identification func-tion that is the purpose of trademark, itdoes not constitute unfair competition;such use is fair because it does not implysponsorship or endorsement by thetrademark holder.

Id. at 308.

The Court distinguished ‘‘nominative’’fair use from ‘‘classic’’ fair use, noting thatif defendant’s use of the trademark re-ferred to something other than the plain-tiff’s product, traditional fair use inquirywould continue to govern.

The court then articulated its own testfor nominative fair use:

[W]here the defendant uses a trademarkto describe the plaintiff’s product, ratherthan its own, we hold that a commercialuser is entitled to a nominative fair usedefense provided he meets the followingthree requirements: First, the productor service in question must be one notreadily identifiable without use of thetrademark; second, only so much of themark or marks may be used as is rea-sonably necessary to identify the prod-uct or service; and third, the user mustdo nothing that would, in conjunctionwith the mark, suggest sponsorship orendorsement by the trademark holder.

Id.

In announcing this new test, New KidsOn The Block rejected traditional trade-mark infringement analysis. It held thatthis test replaces the ‘‘likelihood of confu-sion’’ test for trademark cases where nomi-

native fair use is asserted. See Cairns v.Franklin Mint Co., 292 F.3d 1139, 1151(9th Cir.2002). In a more recent case, theNinth Circuit Court of Appeals has elabo-rated on its reasoning for this shift. Ap-plication of the likelihood of confusion test,which focuses on the similarity of the markused by the plaintiff and defendant, ‘‘wouldlead to the incorrect conclusion that virtu-ally all nominative uses are confusing’’ be-cause ‘‘[w]hen a defendant uses a trade-mark nominally, the trademark will beidentical to the plaintiff’s mark, at least interms of the words in question.’’ PlayboyEnters., 279 F.3d at 801. Thus, the likeli-hood of confusion test as applied in nomi-native fair use cases would disadvantagethe defendant by making confusion an allbut foregone conclusion.

While we agree with the Ninth CircuitCourt of Appeals that a distinct analysis isneeded for nominative fair use cases, wedo not accept the legal basis or advisabilityof supplanting the likelihood of confusiontest entirely. First, we do not see nomi-native fair use as so different from classicfair use as to warrant such different treat-ment. The Ninth Circuit Court of Appealsbelieved that the two types of fair usecould be distinguished on the basis thatnominative fair use makes it clear to con-sumers that the plaintiff, not the defen-dant, is the source of the trademarkedproduct or service, while classic fair usedoes not. Thus, the Ninth Circuit Courtof Appeals believed that a different analy-sis was appropriate for nominative fair useand that it could abandon the need forproof of confusion in these circumstances.New Kids On The Block, 971 F.2d at 307–08.1 Yet, it is clear to us that even a

1. Interestingly, the thinking of the Ninth Cir-cuit Court of Appeals appears to have evolvedin this regard. The court reasoned in NewKids On The Block that nominative use doesnot risk consumer confusion and is ‘‘outside

the strictures of trademark law.’’ 971 F.2d at308. Ten years later, however, it justifiedreplacing the likelihood of confusion test byarguing that the test would be inaccurate inthe nominative use context. Playboy Enters.,


defendant’s nominative use has the poten-tial of confusing consumers with respect toits products or services. Since the defen-dant ultimately uses the plaintiff’s mark ina nominative case in order to describe itsown product or services, Cairns, 292 F.3dat 1151 & n. 8, even an accurate nomi-native use could potentially confuse con-sumers about the plaintiff’s endorsementor sponsorship of the defendant’s productsor services. Thus, we disagree with thefundamental distinction the Ninth CircuitCourt of Appeals drew between classic andnominative fair use.

In addition, the approach of the Court ofAppeals for the Ninth Circuit would re-lieve the plaintiff of the burden of provingthe key element of a trademark infringe-ment case—likelihood of confusion—as aprecondition to a defendant’s even havingto assert and demonstrate its entitlementto a nominative fair use defense. TheSupreme Court in KP Permanent Make–Up clearly established that it was plain-tiff’s burden in a classic fair use case toprove likelihood of confusion. There, theCourt noted the difference between fairuse and other trademark infringementclaims, opining, as we stated above, thatlikelihood of confusion and fair use cancoexist. This does not mean that weshould remove the need for finding confu-sion in the first instance. Instead, oncethe plaintiff proves likelihood of confusion,defendant only had to show that defen-dant’s use, even if confusing, was ‘‘fair.’’ 2

[5] This view finds support not only inthe Supreme Court’s recent opinion, butalso in the relevant statutory framework.The very language of the Lanham Actleads us to conclude that likelihood of con-fusion is an essential indicator of whetheror not trademark infringement has oc-curred. Both §§ 32 and 43(a) of the Lan-ham Act, allegedly violated in this case,forbid use of words or marks in a waywhich is likely to cause confusion as to theorigin, sponsorship, or approval of goodsor services. Surely the plaintiff’s successin its claim must rely on a finding in thisregard. Given this, we decline to read thisrequirement out of a case alleging trade-mark infringement.

[6] We are thus left with the firm con-viction that the burden of proving likeli-hood of confusion should remain with theplaintiff in a trademark infringementcase—including one where the defendantclaims nominative fair use. As detailedbelow, we will devise a modified likelihoodof confusion test to be employed in nomi-native fair use cases that takes into ac-count the concerns expressed by the Courtof Appeals for the Ninth Circuit. Then,we will determine the extent to which wewould adopt the test for nominative fairuse that the Ninth Circuit Court of Ap-peals established in light of our disagree-ment with that court’s view that nomi-native fair use is fundamentally differentfrom classic fair use.

279 F.3d at 801. As explained below, webelieve that neither rationale requires us toabandon the likelihood of confusion test inthe nominative use context, and particularlynot since the Supreme Court’s decision in KPPermanent Make–Up.

2. As described above, the Ninth Circuit Courtof Appeals stated its view in New Kids On TheBlock that the use of another’s mark to de-scribe plaintiff’s own product is one step re-

moved from the true scope of the trademarklaws and the very concept of trademark in-fringement. But this thinking provides all themore reason that a plaintiff in that situationshould be required to prove that confusionexists before putting the defendant to itsproof. If such cases are truly outside of thescope of the trademark laws then a plaintiffwould simply fail to show confusion and thecourt’s inquiry would end.


V. The Proper Analytical Approach forNominative Fair Use Cases

A. Overview

[7, 8] Today we adopt a two-step ap-proach in nominative fair use cases. Theplaintiff must first prove that confusion islikely due to the defendant’s use of plain-tiff’s mark. As we discuss more fully be-low, because our traditional likelihood ofconfusion test does not apply neatly tonominative fair use cases, we suggest elim-inating those factors used to establish con-fusion in other trademark infringementcases that do not ‘‘fit’’ in the nominativeuse context. Once plaintiff has met itsburden of proving that confusion is likely,the burden then shifts to defendant toshow that its nominative use of plaintiff’smark is nonetheless fair. To demonstratefairness, the defendant must satisfy athree-pronged nominative fair use test, de-rived to a great extent from the one articu-lated by the Court of Appeals for theNinth Circuit. Under our fairness test, adefendant must show: (1) that the use ofplaintiff’s mark is necessary to describeboth the plaintiff’s product or service andthe defendant’s product or service; (2)that the defendant uses only so much ofthe plaintiff’s mark as is necessary to de-scribe plaintiff’s product; and (3) that thedefendant’s conduct or language reflect thetrue and accurate relationship betweenplaintiff and defendant’s products or ser-vices.

As an initial matter, we recognize thatour concurring colleague rejects the bifur-cated approach that we now adopt. Heargues instead that the factors we considerunder the fairness test should be incorpo-rated into the likelihood of confusion anal-ysis. In his view, our bifurcated approachplaces a heavy burden on the defendant tonegate confusion and is judicially unman-ageable. However, our approach doesnothing of the kind.

We conclude that the broad based likeli-hood of confusion test our concurring col-league proposes is misplaced for severalreasons. First, it is largely out of syncwith the existing jurisprudence on fair use.The concurrence’s test allows no real pos-sibility of the co-existence of fair use withsome likelihood of confusion, yet this isprecisely what the Supreme Court’s hold-ing in KP Permanent Make–Up specifical-ly contemplates. See KP PermanentMake–Up, 125 S.Ct. at 550 (‘‘[S]ome possi-bility of consumer confusion must be com-patible with fair useTTTT’’). In addition,our concurring colleague rejects the notionthat nominative fair use could be used asan affirmative defense, viewing it insteadas a confusion substitute. Yet, the Su-preme Court clearly views fair use (albeitclassic fair use) as an affirmative defense.Id. at 548–49 (referring to the affirmativedefense of fair use). The concurrence failsto explain why KP Permanent Make–Upshould neither control nor inform our anal-ysis here, choosing instead to ignore theCourt’s dictates in that case as they applyto nominative fair use.3 After that deci-

3. The concurrence relies heavily on a mix ofcase law that was, in our view, largely under-mined by the Supreme Court’s holding in KPPermanent Make–Up. Before the SupremeCourt’s opinion in that case, a finding oflikelihood of confusion sounded the deathknell for a classic fair use defense in somecircuits. See KP Permanent Make–Up, Inc. v.Lasting Impression I, Inc., 328 F.3d 1061,1072 (9th Cir.2003) (‘‘[T]here can be no fair

use if there is a likelihood of confusionTTTT’’);PACCAR, 319 F.3d at 256 (‘‘A finding of likeli-hood of confusion forecloses a [classic] fairuse defense.’’). Likewise, the Courts of Ap-peals for the Fifth and Ninth Circuits stated,prior to KP Permanent Make–Up, that nomi-native fair use is not an infringement so longas there is no likelihood of confusion. Cairns,292 F.3d at 1151; Pebble Beach, 155 F.3d at546. These courts have not considered the


sion, it seems to us that neither classic ornominative fair use should rise and fallbased on a finding of likelihood of confu-sion. Classic fair use and nominative fairuse are different in certain respects, but itis unclear to us why we should ask radical-ly different questions when analyzing adefendant’s ability to refer to a plaintiff’smark in the two contexts. As we havealready stated, in both nominative andclassic fair use cases the defendant usesthe plaintiff’s mark descriptively in a waythat potentially confuses consumers aboutthe relationship between the plaintiff andthe defendant’s product or services. Inthe classic fair use context, the defendantuses the mark to describe its own product,and in the nominative context, the defen-dant references plaintiff’s product in orderto describe its own. The key first inquiryin both situations should be whether thereis a likelihood of confusion.4 The onlyother court to consider the application ofnominative fair use doctrine since KP Per-manent Make–Up has embraced this logic.Ty, Inc. v. Publ’ns Int’l, Ltd., No. 99 C5565, 2005 WL 464688, at *6–8 (N.D.Ill.Feb.25, 2005).

Second, while the concurrence worriesthat a nominative use defendant will beoverly burdened under our bifurcated ap-proach, we believe that his approach isactually more burdensome to such a defen-dant. If the factors for determining fair-ness were incorporated into the likelihoodof confusion test, a plaintiff’s showing ofconfusion might well overwhelm a defen-dant’s showing of fair use. This wouldessentially force a defendant assertingnominative fair use to negate all likelihoodof confusion to succeed, a proposition thatthe Supreme Court rejected in KP Perma-nent Make–Up. Under our approach, thedefendant has no duty to negate confusionas such, but rather must merely show thatits use of the plaintiff’s mark is fair, aburden which, by contrast, is not cumber-some. Thus, it is our view that the bifur-cated approach is ultimately less burden-some to a nominative use defendant thanthe analysis the concurrence proposes.

Finally, we believe that the bifurcatedapproach that we adopt today is moreworkable than a unified confusion/fairnesstest. We leave the now familiar test forlikelihood of confusion largely intact and in

viability of their approach since KP PermanentMake–Up was decided. See Ty, Inc. v. Publ’nsInt’l, Ltd., infra, at *7. We rely on the NinthCircuit precedent insofar as it identifies theneed for a separate test for nominative fairuse, but we reject its view that this test shouldbe a substitute for likelihood of confusion. Anominative use defendant need only provefairness and is not required to negate confu-sion.

4. Judge Fisher points to Prestonettes, Inc. v.Coty, 264 U.S. 359, 44 S.Ct. 350, 68 L.Ed. 731(1924), and G.D. Searle & Co. v. Hudson Phar-maceutical Corp., 715 F.2d 837 (3d Cir.1983),as authority for limiting our inquiry to likeli-hood of confusion. However, even a cursoryreview of those opinions-dating back 80 and20 years, respectively-reveals little in the wayof guidance for an analysis of nominative fairuse. Indeed, no reference to ‘‘fair use,’’ letalone ‘‘nominative fair use,’’ can be found in

either opinion. Both cases seem, instead, tostand for the proposition that clear and truth-ful references to the products of others arepermitted. Thus, the underlying theme ofthose opinions seems to support, rather thanundercut, an analysis that specifically exam-ines whether the use is clear and truthful, asdoes the affirmative defense we posit, andthat the Supreme Court has embraced.

Moreover, our analysis does not disturb theholding of either Prestonettes or Searle. Afterour decision today, courts will continue toinquire into confusion, as they did in both ofthose cases. Since KP Permanent Make–Uphas altered the landscape, however, we be-lieve it is clear that likelihood of confusionneed not, and should not, be solely determina-tive of nominative fair use. Fairness is adistinct concept from confusion, and it shouldbe measured through a distinct inquiry.


the form in which district courts are accus-tomed to applying it. Our test for nomi-native fair use considers distinct factorsthat are readily susceptible to judicial in-quiry. By contrast, the concurrence wouldincorporate several new considerationsinto the already lengthy ten-part test forconfusion and ask district courts to balancea plaintiff’s showing of confusion against adefendant’s showing of fair use. Becauseconfusion and fairness are separate anddistinct concepts that can co-exist, blend-ing them together into one test is, to ourmind, a much less manageable approach.

B. The Proper Test For Likelihoodof Confusion

[9] As we have noted, and as the NinthCircuit in New Kids on the Block alsostated, the ‘‘likelihood of confusion’’ testdoes not lend itself nicely to a nominativefair use fact pattern. The traditional like-lihood of confusion test has been set forthin our case law, specifically in Scott PaperCo. v. Scott’s Liquid Gold, Inc., 589 F.2d1225 (3d Cir.1978) and Interpace Corp. v.Lapp, Inc., 721 F.2d 460, 463 (3d Cir.1983).It is a multi-factor test that assesses thefollowing:

(1) degree of similarity between theowner’s mark and the alleged in-fringing mark;

(2) strength of the owner’s mark;

(3) price of the goods and other factorsindicative of the care and attentionexpected of consumers when makinga purchase;

(4) length of time the defendant hasused the mark without evidence ofactual confusion;

(5) intent of the defendant in adoptingthe mark;

(6) evidence of actual confusion;

(7) whether the goods, though not com-peting, are marketed through the

same channels of trade and adver-tised through the same media;

(8) the extent to which the targets ofthe parties’ sales efforts are thesame;

(9) the relationship of the goods in theminds of consumers because of thesimilarity of function; and

(10) other facts suggesting that the con-suming public might expect the pri-or owner to manufacture a productin the defendant’s market or thathe is likely to expand into that mar-ket.

589 F.2d at 1229.

We have come to call these factors the‘‘Lapp’’ factors. Although they are oftenhelpful in determining whether a certainuse of a mark is likely to confuse consum-ers, we have recognized that ‘‘the LanhamAct does not require that they be followedprecisely so long as the relevant compari-sons suggested by the test are made.’’ A& H Sportswear II, 237 F.3d 198, 207 (3dCir.2000). In the context of a nominativeuse of a mark, such as the one we arepresented with here, certain Lapp factorsare either unworkable or not suited orhelpful as indicators of confusion in thiscontext. That is because, by definition,nominative use involves the use of anoth-er’s trademark in order to describe thetrademark owner’s own product. Further,certain of the Lapp factors applied me-chanically would inevitably point towardslikelihood of confusion where no likelihoodof confusion may actually exist. Thus, wemust tailor the test and measure onlythose factors that are meaningful and pro-bative in the context of nominative fairuse.

Specifically, the first two Lapp factorswould indicate a likelihood of confusion ina case such as this one simply because themark is being employed in a nominativemanner. By way of example, looking at


the similarity of the mark would automati-cally lead to the conclusion that the use islikely to confuse simply because the markis not merely similar, it is identical. Thefirst Lapp factor does not leave any roomfor the consideration of the context of theuse—i.e., that the mark is being used todescribe the plaintiff’s own product.Therefore, it is not appropriate for analy-sis in a nominative use case.

Looking at the strength of CCE’s marksin this case, and in most nominative usecases, would also weigh in favor of a find-ing that the use is likely to confuse. How-ever, defendants in nominative use cases,like LT in this case, feel they need to usethe actual mark to describe the plaintiff’sproduct because of its very strength andwhat it has come to represent. In reality,in many such cases, the use of the namemay be the only way for defendant toeasily and precisely refer to plaintiff’sproduct in a way that will be understoodby consumers. Accordingly, the marks’strength is not really probative of confu-sion here, whereas it would be if defendantwere passing off its goods under a similarmark.

We find that all of the other Lapp fac-tors, while perhaps not appropriate foranalysis in this particular case, could beanalyzed in future nominative use cases,depending on the factual situation. Forexample, in this case, looking at whetherthe goods are marketed through the samechannels or advertised in the same mediawould be a misleading indicator in deter-mining the true likelihood of confusion. Itwould be of little relevance in this casewhether LT marketed its services throughthe same channels as CCE because LT isnot attempting to use CCE’s marks torefer to LT’s own services, but rather isusing the marks to refer to CCE’s ser-vices. Therefore, it would almost be ex-pected that LT and most other defendants

in a nominative use case would marketthemselves through a media in which themarks to which they are referring wouldbe easily recognized and have meaning orrelevance, namely channels similar tothose used by plaintiff. However, theremay be certain situations, such as thatencountered in New Kids (where newspa-pers used the New Kids’ mark to inquireof readers, for a price, their feelings aboutthe group), where the channels of market-ing are so dissimilar that evidence as tothis factor could mitigate against a findingof likelihood of confusion. Similarly, inlooking at whether the targets of the par-ties’ marketing efforts are the same, onecould expect a fair use defendant to bereaching out to a group of consumers whoare likely to recognize the mark. The factthat LT markets its services to real estateconsumers makes sense. These are thevery people who are likely to recognizeand appreciate CCE’s marks and be ableto evaluate the benefit that CCE’s pur-ported association with LT could bring tothem in the search for a home. This is notlikely to confuse in the same way as wouldbe the case if the defendant is passing offplaintiff’s goods as its own or using asimilar mark to that of the plaintiff, butthis factor should not be completely elimi-nated from a district court’s arsenal inevaluating likelihood of confusion in a nom-inative use case.

[10] In the context of the facts of thecase before us, we will turn our focus tothose Lapp factors that appear most rele-vant in assessing likely confusion. Theseinclude:

(1) the price of the goods and otherfactors indicative of the care andattention expected of consumerswhen making a purchase;

(2) the length of time the defendant hasused the mark without evidence ofactual confusion;


(3) the intent of the defendant in adopt-ing the mark; and

(4) the evidence of actual confusion.In focusing on these factors, the court willbe better able to assess whether consum-ers are likely to be confused by the use notbecause of its nominative nature, but rath-er because of the manner in which themark is being nominatively employed. Ofcourse the determination of which factorsare relevant and probative in a given factu-al setting should be made in the first in-stance by the District Court, but here weprovide our view on this issue as guidance,lest this matter again should come before apanel of this Court on the issue of whetherthe District Court examined the appropri-ate factors.

Viewed in the context of the presentcase, it is apparent why these factors areappropriate: they analyze the likelihoodthat a consumer will be confused as to therelationship or affiliation between LT andCCE, the heart of the nominative fair usesituation. Focusing on the care and atten-tion expected of consumers when making apurchase or using a service allows thecourt to understand the true risk that con-sumers may be confused merely because oftheir own inattention. Moreover, focusingon evidence of actual confusion (factors 2and 4) will allow the District Court to trulyunderstand whether this is merely a theo-retical or hypothetical fear of the plaintiffor whether there is real danger that con-sumers are likely to be confused.

Lastly, analyzing the intent of the defen-dant in using the mark will allow the courtto understand the defendant’s reason forutilizing the mark in the manner that itdid. That is, if the court finds that thedefendant made use of the plaintiff’s markwith the very purpose of causing consum-ers to think the plaintiff endorses or spon-

sors plaintiff’s good or service, then thelikelihood that consumers will be confusedas to endorsement/affiliation is greater.Whereas the traditional ‘‘intent’’ pronglooks at intent to adopt a similar mark,here the key inquiry is whether the markis being used so as to convey a connectionbetween the parties that may not exist.

We hold today that the burden of prov-ing likelihood of confusion, even in a nomi-native use case, should remain with theplaintiff and that these four factors are theessence of the inquiry, although othersmay prove useful in certain contexts.Here, the District Court did place thisburden on CCE, but the District Courtalso analyzed certain Lapp factors that wehave concluded are inappropriate in thecontext of this nominative fair use case.Further, its application of certain of thefour relevant factors was somewhat impre-cise.

[11] We will remand for the DistrictCourt to inquire as to whether confusionwas likely in light of our clarification ofwhich factors are applicable in this caseand the following discussion of the way inwhich the four relevant factors should beapplied.5

Factor 1: The Price of the Goods andOther Factors Indicative of the Careand Attention Expected of ConsumersWhen Making a Purchase

The District Court noted that ‘‘we aretalking about real estate purchases whichin reference to most consumers is the mostimportant purchase both at the time and,in most people’s instances, in their entirelives.’’ (Trans. of Oral Opinion at JA735.)The District Court was correct to acknowl-edge and consider the importance of a real

5. We review a district court’s conclusion as tolikelihood of confusion for clear error. A &

H Sportswear II, 237 F.3d 198, 237 (3d Cir.2000).


estate purchase to the average consumer.One would expect that consumers going toLT website’s ‘‘Find a Realtor’’ sectionwould be very careful in learning about thevarious realtors in LT’s network. The de-gree of care weighs against a finding oflikely confusion.

Factor 2: The Length of Time the De-fendant has used the Mark without Evi-dence of Actual Confusion

The District Court conflated this factorwith the defendant’s laches defense and,therefore, failed to weigh in LT’s favor thefact that, according to the record, CCE’smarks had been used for at least five yearswith little evidence of actual consumer con-fusion.

Factor 3: The Intent of the Defendantin Adopting the Mark

[12] The District Court appears tohave inferred improper intent on the partof LT from its very use of CCE’s marks.This finding was erroneous without furtherevidentiary support. The relevant ques-tion in this context is not whether thedefendant intended to use the plaintiff’smark, which it always has in a fair usecase, but whether it used the mark withthe intent to confuse the public as to therelationship between the defendant andthe plaintiff. Use of the mark alone is notsufficiently probative of such intent.6 Theplaintiff must show that the defendant in-

tended the public to believe that the plain-tiff endorsed or somehow supported itsproducts or services.

There are myriad factors to which aplaintiff may point to demonstrate that adefendant intended to confuse the publicas to its relationship with the plaintiff.For example, courts in traditional trade-mark infringement cases have considered adefendant’s persistence in adopting a markdespite being warned of potential confu-sion, KOS Pharms., Inc. v. Andrx Corp.,369 F.3d 700, 721 (3d Cir.2004), and evi-dence of a defendant’s bad faith, A & HSportswear II, 237 F.3d at 226. On re-mand, the District Court should inquire asto whether any analogous facts or evidenceof record are present here. If so, theCourt may weigh this in favor of a findingof likely confusion.7

Factor 4: The Evidence of Actual Con-fusion.

In this case, CCE presented no evidenceof actual confusion before the DistrictCourt. In fact, as the District Court not-ed, from the Court’s first meeting with theparties, they had agreed that they wouldoffer no evidence of actual confusion to theCourt. Therefore, this factor should clear-ly weigh in favor of LT.

We will remand the case to the DistrictCourt for proper application of these fac-

6. This principle emerges from our analogousholdings in traditional trademark infringe-ment cases, in which we have noted that ‘‘adefendant’s mere intent to copy, withoutmore, is not sufficiently probative of the de-fendant’s success in causing confusion toweigh such a finding in the plaintiff’s favor.’’A & H Sportswear II, 237 F.3d at 225–26.Rather, the plaintiff should make some show-ing that the defendant adopted its mark withthe intent to confuse or deceive the public.Id.; see also Checkpoint Sys. Inc. v. CheckPoint Software Techs., Inc., 269 F.3d 270, 286(3d Cir.2001) (noting that ‘‘a party’s inten-tional use of another party’s mark to cause

confusion’’ weighs in favor of finding likeli-hood of confusion) (emphasis added).

7. The concurrence argues that our nomi-native fair use defense is merely this sameintent inquiry dressed in different clothing.The tests for confusion and fairness are notduplicative, however, because a plaintiff canalso establish likelihood of confusion throughseven other Lapp factors besides intent. It isthe circumstance in which a court does notfind bad intent but does find confusion that anominative fair use defense will be most use-ful.


tors. In determining whether confusion islikely, as we noted in A & H SportswearII, 237 F.3d 198, 215 (3d Cir.2000), ‘‘differ-ent factors may be accorded differentweights depending on the particular factu-al setting.’’ If CCE is able to meet itsburden of showing likelihood of confusionon remand, the burden of proof will thenshift to LT to show that the use is anominative fair use.

C. The Affirmative Defense of Nomi-native Fair Use

Under the nominative fair use testadopted by the Court of Appeals for theNinth Circuit, a defendant must prove: (1)that the product or service in question isone not readily identifiable without use ofthe trademark; (2) that only so much ofthe mark or marks is used as is reasonablynecessary to identify the product or ser-vice; and (3) that the user did nothing thatwould, in conjunction with the mark, sug-gest sponsorship or endorsement by thetrademark holder. New Kids on theBlock, 971 F.2d at 308. In the NinthCircuit, if these elements are proven, theuse is ‘‘fair’’ and defendant will prevail.Further, this nominative fair use test, asdiscussed above, replaces the likelihood ofconfusion test in the Court of Appeals forthe Ninth Circuit. We must decide theextent to which we should adopt this testas our own, mindful that we will employ itas an affirmative defense to be proven bydefendant after likelihood of confusion hasbeen demonstrated by the plaintiff.

We are tempted to use the three-pronged Ninth Circuit test outright, as ithas withstood the test of time, has beentinkered with in no less than seven opin-

ions.8 Moreover, the Ninth Circuit Courtof Appeals is the only Court of Appeals tohave specifically crafted its own test fornominative fair use. Indeed, the SupremeCourt decision in KP Permanent Make–Up involved a reversal of the most recentNinth Circuit opinion on the issue of ‘‘fairuse.’’ As prolific as the jurisprudence ofthe Ninth Circuit Court of Appeals hasbeen in this area, the Supreme Court’srejection of its view of how the fair usedefense works in a classic fair use caseemboldens us to examine the elements ofthe Ninth Circuit Court of Appeals’ nomi-native fair use test a bit more closely.

In so doing, we conclude that the test aswritten suffers from a lack of clarity. Thisis evident in the contortions that the NinthCircuit Court of Appeals itself has gonethrough in applying it, the confusion thatthe District Court here encountered in itsapplication, and in our conviction that amodified inquiry would aid in reaching theright result. We will adjust the test toinclude a slightly different set of consider-ations:

1. Is the use of plaintiff’s mark neces-sary to describe (1) plaintiff’s product orservice and (2) defendant’s product or ser-vice?

2. Is only so much of the plaintiff’smark used as is necessary to describeplaintiff’s products or services?

3. Does the defendant’s conduct or lan-guage reflect the true and accurate rela-tionship between plaintiff and defendant’sproducts or services?

The following discussion explains howthese factors should be applied.

8. See, The New Kids on the Block v. GannettSatellite Information Network, Inc., 971 F.2d302 (9th Cir.1992); Kareem Abdul–Jabbar v.General Motors Corp., 85 F.3d 407 (9th Cir.1996); Cairns v. Franklin Mint Co., 292 F.3d1139 (9th Cir.2002); Playboy Enterprises, Inc.

v. Welles, 279 F.3d 796 (9th Cir.2002); Broth-er Records, Inc. v. Jardine, 318 F.3d 900 (9thCir.2003); Horphag Research Ltd. v. Pellegri-ni, 337 F.3d 1036 (9th Cir.2003); PlayboyEnterprises, Inc. v. Netscape CommunicationsCorp., 354 F.3d 1020 (9th Cir.2004).


1. First Prong

The first element of the Ninth CircuitCourt of Appeals’ test involved an inquiryonly into the necessity of using plaintiff’strademark to describe plaintiff’s product.Here, the District Court instead—probablymistakenly—examined the necessity of theuse of plaintiff’s trademark in order todescribe defendant’s product. See Cairnsv. Franklin Mint Co., 292 F.3d 1139, 1153(9th Cir.2002). The first prong of NewKids On The Block is at best confusingand at worst incomplete. While it shouldbe asked whether plaintiff’s product needsto be described by reference to its mark,should it not also be examined whetherdefendant’s use of it, at all, is necessary toaccurately describe what defendant doesor sells, or whether its reference to plain-tiff’s mark is actually gratuitous? TheDistrict Court’s inquiry into the latter as-pect was not called for under the NewKids On The Block test, but it actuallyseems entirely appropriate.

The focus on the necessity of the markin order for defendant to describe plain-tiff’s product makes sense in the context ofnominative fair use, where the plaintiff’smark is being used because it identifies theplaintiff’s product. We further note thatthe court need not find that the use of themark is indispensable in order to find thisfactor fulfilled. For, as we have statedbefore, ‘‘[t]he Lanham Act does not compela competitor to resort to second-best com-munication.’’ G.D. Searle & Co., 715 F.2d837 at 842. Furthermore, as the NinthCircuit has observed:

it is often virtually impossible to refer toa particular product for purposes ofcomparison, criticism, point of referenceor any other such purpose without usingthe mark. For example, reference to a

large automobile manufacturer based inMichigan would not differentiate amongthe Big Three; reference to a largeJapanese manufacturer of home elec-tronics would narrow the field to a dozenor more companies. Much useful socialand commercial discourse would be allbut impossible if speakers were underthreat of an infringement lawsuit everytime they made reference to a person,company or product by using its trade-mark.

New Kids on the Block, 971 F.2d 302 at307. Therefore, the court need only besatisfied that the identification by the de-fendant of plaintiff’s product or servicewould be rendered significantly more diffi-cult without use of the mark.

Additionally, we believe that it is impor-tant for a court to understand how neces-sary the use of the mark is to the identifi-cation of defendant’s product. That is, themore dependent the ready identification ofdefendant’s product is on the descriptionof plaintiff’s product through the employ-ment of plaintiff’s mark, the more likely itis that the use is a fair one.9

In this case, the District Court looked atthis test only in regard to defendant’sproduct. On remand the Court shouldadditionally focus on whether requiring LTto describe plaintiff CCE’s services with-out using its mark is a forced reversion tosecond-best communications. Consider-ations such as the simplicity of descriptionand the likelihood that consumers will un-derstand a given reference to plaintiff’sservices without use of the mark are ap-propriate to this analysis. See Cairns, 292F.3d at 1153.

2. Second Prong

Here again, the New Kids On The Blocktest focuses on the ‘‘amount’’ of plaintiff’s

9. Although the test crafted by the Ninth Cir-cuit Court of Appeals did not include thisaspect under the first prong, as such, none-

theless, it does allude to it as part of thesecond prong, as we discuss below.


mark that is used, asking whether only somuch is used as is necessary to describeplaintiff’s product. That focus on limitingthe quantum use of the attributes of plain-tiff’s marks’ is appropriate, for it is the useof plaintiff’s marks that concerns us. Yet,the case law of the Ninth Circuit Court ofAppeals has expanded on this prong toinquire, at this stage, about the defen-dant’s need to use of the mark, stating:

[W]hat is ‘reasonably necessary to iden-tify the plaintiff’s product differs from caseto caseTTTT Where, as in the presentcase, the description of the defendant’sproduct depends on the description of theplaintiff’s product, more use of the plain-tiff’s trademark is ‘reasonably necessary toidentify the plaintiff’s product’ than incases where the description of the defen-dant’s product does not depend on thedescription of the plaintiff’s product.

Cairns, 292 F.3d at 1154. Since we havepositioned the assessment of the defen-dant’s need to use plaintiff’s mark as partof the first inquiry, we eliminate the confu-sion inherent in inquiring into ‘‘need’’ atthe second step. Thus, under our ap-proach, the second prong tests only wheth-er the quantum of the plaintiff’s markused by the defendant was appropriate.

In analyzing this factor, the DistrictCourt essentially predetermined the out-come of the second prong by its finding asto the first prong. The District Courtfound that because the use of CCE’smarks was not necessary to identify LT’sservices as a whole under the first prong,the use could not possibly be only so muchas was necessary. But the proper focusunder this prong is on whether only somuch of plaintiff’s mark as is reasonablynecessary to identify plaintiff’s product orservice has been used by defendant. Con-sideration should be given at this stage tothe manner in which the mark is por-trayed. For example, did the defendant

use plaintiff’s distinctive lettering when us-ing the plaintiff’s mark or did the defen-dant, as in this case, simply use blockletters to spell out plaintiffs’ names?

In his concurring opinion, Judge Fisherposits that both Prongs One and Two ofour modified nominative fair use test are,and should be, subsumed in the likelihoodof confusion analysis. He suggests thatthe necessity of using the mark (Prong 1)and the reasonableness of that use (Prong2) are just another way of examining thedefendant’s intent in using the mark. Thisis a mischaracterization of both the intentfactor under Lapp, as described above, andthe first two prongs of the nominative usedefense we espouse. Even if it is foundthat a defendant’s intent is improper andits use of a mark could cause confusion,the defendant could conceivably prove thateven its potentially confusing use is fairbecause it is reasonably limited and neces-sary under the first two prongs of ournominative fair use test. These contoursof fairness are not recognized by our con-curring colleague, but they are fundamen-tal to the concept of fair use according tothe Supreme Court’s recent decision in KPPermanent Make–Up.

3. Third Prong

The New Kids test at this stage askswhether the user did anything that would,in conjunction with the mark, suggestsponsorship or endorsement by the trade-mark holder. However, we believe theappropriate question should be a bitbroader: does the defendant’s conduct orlanguage reflect the true and accurate re-lationship between plaintiff and defen-dant’s products or services? We believethat the Ninth Circuit Court of Appeals’focus on whether the user’s conduct im-plies endorsement may not truly reflectwhether the use is fair because sometimesthe plaintiff’s relationship with defendant


may be one of endorsement, but the natureof the endorsement as reflected by defen-dant’s employment of plaintiff’s marks maynot be accurate. A defendant’s purposefulportrayal of plaintiff’s endorsement of itsproduct through defendant’s conduct orlanguage does not necessarily render theuse unfair, as long as the depiction of theendorsement is accurate. In addition, ourversion suggests that we can consider thedefendant’s failure to state or explain someaspect of the relationship, whereas theNew Kids version focuses on affirmativeacts, i.e., what the defendant did to sug-gest sponsorship.

[13] In this case, the District Courtconcluded that an ordinary consumerwould look at LT’s website and concludethat there was some affiliation between LTand the enumerated companies. Further,the Court inferred from the mere presenceof the marks an intent on the part of LT toconvey endorsement or affiliation. On re-mand, the District Court should specifical-ly determine in analyzing this prongwhether the portrayal of the relationshipwas accurate, and what more the defen-dant could have done to prevent an im-proper inference regarding the relation-ship. The mere presence or use of themark does not suggest unfairness underthis prong. Here, LT added a disclaimer,the significance of which the District Courtdownplayed, stating that ‘‘LendingTree isnot sponsored by or affiliated with theparent franchisor companies of any of theparticipating members of its network.’’(Text from www.LendingTree.com atJA687.) Far from unimportant, such adisclaimer must be considered in determin-ing whether the alleged infringer accurate-ly portrayed the relationship that existedbetween plaintiff and defendant. SeePlayboy Enters., 279 F.3d at 803 (findingthat the fact that defendant’s website in-cluded a clear disclaimer of any connection

to plaintiff satisfied this prong of the NinthCircuit Court of Appeals’ test for nomi-native fair use). Therefore, the DistrictCourt on remand should look at the pre-cise way in which what the defendant said,or did not say, other than the mere pres-ence of the mark on the website, may haveinaccurately implied endorsement or spon-sorship by CCE. The District Court shouldconsider whether the disclaimer was anaffirmative action by LT that effectivelynegated an inaccurate implication of spon-sorship or endorsement by CCE.

The Ninth Circuit’s test for nominativefair use does not explicitly include accura-cy within the analysis, but the SupremeCourt has recognized that ‘‘[a]ccuracy ofcourse has to be a consideration in assess-ing fair use.’’ KP Permanent Make–Up,Inc., 125 S.Ct. at 551. In examining theconduct of the defendant to determinewhether the defendant has done anythingto affirmatively cause consumer confusion,it is only reasonable to consider the preci-sion with which the defendant has de-scribed its relationship with plaintiff. Inthis case, the evaluation of accuracy wouldnecessarily include consideration of LT’scharacterization of the nature and extentof its relationship with CCE and itsagents. This would include the DistrictCourt’s consideration of LT’s reference toits affiliation with CCE brokers in general,rather than referencing each more accu-rately by their ‘‘d/b/a/’’ title and whetherthis rendered the use inaccurate or wassomehow misleading as to any endorse-ment or relationship.

V. Conclusion

In sum, we hold today that the Lapptest for likelihood of confusion still has animportant place in a trademark infringe-ment case in which the defendant assertsthe nominative fair use defense. In thiscase, the test should focus on the four


relevant factors: (1) the price of the goodsand other factors indicative of the care andattention expected of consumers whenmaking a purchase; (2) the length of timethe defendant has used the mark withoutevidence of actual confusion; (3) the intentof the defendant in adopting the mark;and (4) the evidence of actual confusion.

Once plaintiff has met its burden ofproving that confusion is likely, the burdenthen shifts to defendant to show that itsnominative use of plaintiff’s marks is none-theless fair. In this Circuit, we have todayadopted a test for nominative fair use inwhich a court will pose three questions:(1) Is the use of the plaintiff’s mark neces-sary to describe both plaintiff’s product orservice and defendant’s product or service?(2) Is only so much of the plaintiff’s markused as is necessary to describe plaintiff’sproducts or services? (3) Does the defen-dant’s conduct or language reflect the trueand accurate relationship between plaintiffand defendant’s products or services? Ifeach of these questions can be answered inthe affirmative, the use will be considereda fair one, regardless of whether likelihoodof confusion exists.

We adopt a bifurcated approach thattests for confusion and fairness in separateinquiries in order to distribute the burdenof proof appropriately between the partiesat each stage of the analysis. The defen-dant has no burden to show fairness untilthe plaintiff first shows confusion. Fur-thermore, by properly treating nominativefair use as an affirmative defense, ourapproach allows for the possibility that adistrict court could find a certain level ofconfusion, but still ultimately determinethe use to be fair. By contrast, a unifiedlikelihood of confusion test would require adefendant to negate likelihood of confusionby undercutting the Lapp factors. Be-cause the Supreme Court explicitly reject-

ed such a proposition in KP PermanentMake–Up, we decline to adopt it.

We will therefore REVERSE the orderof the District Court and REMAND forfurther proceedings consistent with thisopinion.

FISHER, Circuit Judge.

As the majority correctly concludes,proper nominative use is permissible use,leaving the critical question of how toframe the nominative use analysis. I con-cur with the majority’s ‘‘firm convictionthat the burden of proving likelihood ofconfusion should remain with the plaintiffin a trademark infringement case.’’ Maj.Op. at 221. I also concur in its conclusionthat this Court’s Lapp factors,10 where rel-evant, must be used in the analysis. Itherefore join in the majority’s judgmentremanding the case for further proceed-ings.

I depart from majority, however, in itsadoption of a bifurcated analysis that looksfirst to a truncated Lapp-factor test, fol-lowed by an affirmative defense of ‘‘nomi-native fair use.’’ Despite professing a‘‘firm conviction’’ that the burden of prov-ing likely confusion remains on plaintiffs,the majority formulates an affirmative de-fense that shifts to defendant the burdenof negating confusion. In so doing, themajority flouts binding caselaw holdingthat proper nominative use is use that isnot likely to confuse, and that a plaintiffalone bears the burden of establishing like-ly confusion. Moreover, to the extent themajority places any burden on plaintiffs atall, it is so watered-down that plaintiffsmight prove likely confusion on one Lappfactor alone. The majority’s bifurcatedtest is also judicially unmanageable be-cause it requires courts to address identi-cal factors on both sides of the equation.

10. Interpace Corp. v. Lapp, Inc., 721 F.2d 460 (3d Cir.1983).


Accordingly, I dissent. I first reviewnominative use cases and conclude thatprecedent, binding and otherwise, showsthat the question of nominative use is aquestion of likely confusion and not anaffirmative defense. Second, I address themajority’s bifurcated analysis, and con-clude that it runs afoul of the Lanham Actand binding precedent because it places ondefendant the burden of negating likelyconfusion. I also conclude that the test isjudicially unmanageable because it re-quires courts to examine identical likeli-hood of confusion factors on both sides ofthe analysis. Finally, I propose that innominative use cases, courts apply a modi-fied likelihood of confusion inquiry basedon Lapp factors two through ten, with theburden resting where it must, on plaintiffs.Because proper nominative use is use thatis not likely to confuse, it is the majority’s‘‘nominative fair use’’ test that is unfair,because it places the burden where itshould never be, on the defendant.

I.

Despite the majority’s desire to ‘‘clarifythe proper analysis,’’ Maj. Op. at 219, itsapproach significantly increases the uncer-tainty and—to use a term appropriate tothe present context—confusion in this im-portant area of the law. In light of thisuncertainty, and in light of the SupremeCourt’s holding in KP Permanent Make–Up, Inc. v. Lasting Impression I, Inc., 543U.S. 111, 125 S.Ct. 542, 160 L.Ed.2d 440(2004), that a defendant bears no burdento negate confusion, it is vital to examinethe nature of nominative use to determinewhether it is an affirmative defense, orinstead an established way to test for like-lihood of confusion. Because it is the lat-ter, the burden lies with plaintiffs and abifurcated analysis is improper.

The majority, stating that the Ninth Cir-cuit is the only Court of Appeals to craft atest for nominative use, mildly rewordsthat test and recasts it as an affirmativedefense. As discussed in the subsectionsbelow, this approach is unprecedented, un-wise, and impermissible. First, it is un-precedented. The Ninth Circuit’s deci-sions make it clear that the test is nothingmore than a likelihood of confusion substi-tute, making it an inappropriate foundationfor an affirmative defense. Second, it isunwise. Other Courts of Appeals haveexercised great skepticism towards theNinth Circuit’s test and none have adoptedit outright, calling into serious question thewisdom of using it as a analytic model.Finally, the majority’s approach is imper-missible. Binding precedent from the Su-preme Court and this Court requires thatnominative use be addressed through alikelihood of confusion analysis and thatthe burden remain on plaintiffs.

A.

The erroneous major premise underly-ing the majority’s conclusion is the as-sumption that nominative use must be ana-lyzed through an affirmative defense.This assumption begins with the majority’smisuse of the test enunciated by the NinthCircuit in New Kids on the Block v. NewsAmerica Publishing, Inc., 971 F.2d 302(9th Cir.1992). As stated by the majority:‘‘We must decide the extent to which weshould adopt [the Ninth Circuit’s] test asour own, mindful that we will employ it asan affirmative defense to be proven bydefendant after likelihood of confusion hasbeen demonstrated by the plaintiff.’’ Maj.Op. at 228 (emphasis added). However,the Ninth Circuit has made it crystal-clearthat its test is a substitute for its normalmulti-factor likelihood of confusion test, tobe used in cases involving purported nomi-


native use.11 The Ninth Circuit has alsomade it clear that this test places on thedefendant the burden of proving no likeli-hood of confusion. Because KP Perma-nent makes it clear that it is improper toplace a burden of proving nonconfusion onthe defendant, the majority grievously errsby treating the Ninth Circuit test as anaffirmative defense.

As noted by New Kids, classic statutoryfair use applies ‘‘where the defendant hasused the plaintiff’s mark to describe thedefendant’s own product.’’ 971 F.2d at308 (emphasis in original). Thus, if BESTPRICE were used as a mark, third partiesmight still fairly use ‘‘best price’’ in itsoriginal descriptive sense to refer to theirown goods or services, such as ‘‘We offerthe best price on breakfast cereal.’’ Incontrast, nominative use arises ‘‘where thedefendant uses a trademark to describethe plaintiff’s product, rather than itsown.’’ New Kids, 971 F.2d at 308. InNew Kids, one of the defendants, a news-paper, had printed a poll saying ‘‘NewKids on the Block are pop’s hottest group.Which of the five is your fave? Or arethey a turn off?’’ Id. at 304. Anotherdefendant newspaper, under a picture ofone of the New Kids, asked, ‘‘Which of theNew Kids on the Block would you mostlike to move next door?’’ Id. The Courtheld:

a commercial user is entitled to a nomi-native fair use defense provided hemeets the following three requirements:First, the product or service in questionmust be one not readily identifiablewithout use of the trademark; second,only so much of the mark or marks maybe used as is reasonably necessary toidentify the product or service; and

third, the user must do nothing thatwould, in conjunction with the mark,suggest sponsorship or endorsement bythe trademark holder.

Id. at 308 (note omitted). The first re-quirement was met because ‘‘[i]t is nomore reasonably possible, however, to re-fer to the New Kids as an entity than it isto refer to the Chicago Bulls, Volkswagensor the Boston Marathon without using thetrademark.’’ Id. The second requirementwas met because defendants ‘‘do not usethe New Kids’ distinctive logo or anythingelse that isn’t needed to make the an-nouncements intelligible to readers.’’ Id.The third element was met as well because‘‘nothing in the announcements suggestsjoint sponsorship or endorsement by theNew Kids.’’ Id. In particular, one of theads asked ‘‘whether the New Kids mightbe ‘a turn off.’ ’’ Id. at 309.

In subsequent cases, the Ninth Circuithas emphasized that the test is a substi-tute for the normal multifactor likelihoodof confusion analysis—‘‘nominative fair useanalysis TTT replaces the likelihood of cus-tomer confusion analysis.’’ Cairns v.Franklin Mint Co., 292 F.3d 1139, 1151(9th Cir.2002) (emphasis in original); seealso KP Permanent Make–Up, Inc. v.Lasting Impression I, Inc., 328 F.3d 1061,1072 (9th Cir.2003) (‘‘When analyzing nom-inative fair use, it is not necessary to ad-dress likelihood of confusion because thenominative fair use analysis replaces thelikelihood of confusion analysis.’’ (empha-sis added)), rev’d on other gds., 543 U.S.111, 125 S.Ct. 542, 160 L.Ed.2d 440 (2004).

Fueling the majority’s mistaken use ofthe Ninth Circuit’s test as an affirmativedefense is the fact that the Ninth Circuitconfusingly refers to its own test at times

11. Like our Court and other Courts of Ap-peals, the Ninth Circuit uses a multi-factortest for likelihood of confusion that looks to,inter alia, the similarity of the marks, strength

of the plaintiff’s mark, and other factors. SeeAMF Inc. v. Sleekcraft Boats, 599 F.2d 341(9th Cir.1979).


as a defense, such as in New Kids wherethe Court called the test ‘‘a nominative fairuse defense.’’ 971 F.2d at 308. But it isclear that the Ninth Circuit’s test is not anaffirmative defense in the sense the major-ity envisions, as a defense to be assertedonce a plaintiff has established a likelihoodof confusion. This is because the NinthCircuit’s so-called ‘‘defense’’ is itself a like-lihood of confusion test. In Brother Rec-ords, Inc. v. Jardine, the Court acknowl-edged charges of ‘‘ ‘confusion in the caselaw’ ’’ as to whether ‘‘nominative fair use’’was really a distinct defense or instead‘‘ ‘merely one type of use which is notlikely to cause confusion.’ ’’ 318 F.3d 900,908–09 n. 5 (9th Cir.) (quoting 2 J. ThomasMcCarthy, McCarthy on Trademarks andUnfair Competition § 11:47 (4th ed.2001)),cert. denied, 540 U.S. 824, 124 S.Ct. 155,157 L.Ed.2d 45 (2003). In response, theCourt admitted that its test replaced thenormal multi-factor likelihood-of-confusionanalysis, with one significant difference: it‘‘shifts to the defendant the burden ofproving no likelihood of confusion.’’ Id. at909 n. 5.

Thus, it is clear that the Ninth Circuittreats its ‘‘defense’’ as a substitute fortesting likelihood of confusion. It is alsoclear that it places the burden of negatingconfusion on the defendant, which underKP Permanent, is now impermissible. InKP Permanent, the Court held that a de-fendant at no time possesses the burden ofnegating confusion, noting, inter alia, ‘‘theincoherence of placing a burden to shownonconfusion on a defendant,’’ 125 S.Ct. at549, and that ‘‘it takes a long stretch toclaim that a defense of fair use entails anyburden to negate confusion.’’ Id. at 548.As the Supreme Court has made it clearthat the burden of showing likely confusionbelongs to plaintiffs alone, it is improper totwist the nominative use inquiry into anaffirmative defense.

B.

The majority suggests that refusing toadopt its approach would be ‘‘largely out ofsync with the existing jurisprudence onfair use.’’ Maj. Op. at 222. But consider-ing the majority’s concessions that theNinth Circuit ‘‘is the only Court of Appealsto have specifically crafted its own test fornominative fair use,’’ id. at 228, that ‘‘[f]ewother courts have spoken on the preciseissue,’’ id. at 218, and that there is a‘‘paucity of the case law on this subjectfrom every [other] court’’ on nominativeuse, id. at 219, the majority cannot point toa set of existing jurisprudence from whichto stray.

If anything, to the extent there is dis-cernable jurisprudence regarding theNinth Circuit’s approach, the trend isheavily against adopting its test. TheSixth Circuit, faced with the option ofadopting the Ninth Circuit’s test, held‘‘[t]his circuit has never followed the nomi-native fair use analysis, always having ap-plied [our own multi-factor likelihood ofconfusion] test. We are not inclined toadopt the Ninth Circuit’s analysis here.’’PACCAR Inc. v. TeleScan Technologies,L.L.C., 319 F.3d 243, 256 (6th Cir.2003).The Fifth Circuit applied the test only inpart, declining to impose the first elementof the test. See Pebble Beach Co. v. Tour18 I Ltd., 155 F.3d 526, 546 n. 13 (5thCir.1998) (refusing to adopt the first ele-ment because it would always be satisfiedin disputes like the comparative advertis-ing case before the Court, and declining toopine on whether the requirement shouldbe imposed outside the comparative adver-tising context). Pebble Beach expresslyrecognized the Ninth Circuit’s ‘‘defense’’for what it was, a likelihood of confusiontest, stating that ‘‘[w]hile a claim that theuse was to identify the markholder’s goodsor services is analogous to the statutory


fair-use defense, it is in actuality a claimthat the use is noninfringing and thuscreates no likelihood of confusion.’’ Id. at545 (emphasis added).12

Other authorities have noted the defi-ciencies of the Ninth Circuit’s approach.A District Court in Maryland has criticizedall three elements, noting that the first andthird prongs are nothing more than ‘‘arestatement of two basic principles oftrademark law,’’ first, that use of a mark‘‘is not prohibited if the use is intendedmerely to refer to the holder of the mark,’’and second, that such reference ‘‘is permis-sible provided it is not likely to causeconfusion.’’ Nat’l Federation of Blind,Inc. v. Loompanics Enterp., Inc., 936F.Supp. 1232, 1241 (D.Md.1996). Similar-ly, the second prong ‘‘appear[ed] to derivefrom a concern that confusion as to affilia-tion may result if the defendant’s use ofthe plaintiff’s mark exceeds its legitimatereferential purpose.’’ Id. The test hasbeen additionally and properly criticized as‘‘cloud[ing] the issue of who carries theburden to prove likelihood of confusion.’’Derek J. Westberg, Note, New Kids onthe Block v. News America Publishing,Inc.: New Nominative Use Defense In-creases the Likelihood of Confusion Sur-rounding the Fair Use Defense to Trade-mark Infringement, 24 Golden GateU.L.Rev. 685, 709 (1994). The authornotes that the third element requires ‘‘thatthe user must do nothing that would, inconjunction with the mark, suggest spon-sorship or endorsement of the trademarkholder,’’ but wryly notes that this inquiry‘‘must be satisfied as part of the plaintiff’s

prima facie case of infringement.’’ Id. at705. I agree, and more importantly, sodoes the Supreme Court.13

C.

The irony of the majority’s well-intendedbut erroneous ‘‘nominative fair use’’ de-fense is that the majority felt the need tocraft a defense in the first place. Bindingcaselaw from the Supreme Court and thisCourt have analyzed nominative use caseswithout resorting to a separate test andwithout placing the burden of negatingconfusion on defendants. I do not see howthe majority’s approach can be reconciledwith these cases.

In Prestonettes, Inc. v. Coty, the Su-preme Court addressed a case involvingwhat we would today refer to as nomi-native use. 264 U.S. 359, 44 S.Ct. 350, 68L.Ed. 731 (1924). The plaintiff, owner ofthe marks COTY and L’ORIGAN, soldperfume and powder. The defendantbought genuine COTY and L’ORIGANproducts and repackaged them. Theplaintiff sought an injunction to preventthe defendant from any use of the plain-tiff’s marks on the repackaged goods. TheDistrict Court had permitted the defen-dant to sell the plaintiff’s repackagedgoods, with a disclaimer saying that defen-dant was not connected with plaintiff andthat the perfume and powder were respec-tively ‘‘independently rebottled’’ and ‘‘inde-pendently compounded,’’ with every wordin ‘‘letters of the same size, color, type,and general distinctiveness.’’ Id. at 367, 44S.Ct. 350. Writing for the Court, Justice

12. Even the majority admits ‘‘it seems thatonly the Second, Fifth, and Sixth Circuitshave referenced the nominative fair use de-fense by name and even on these occasionshave done so only to refer to what districtcourts had done with the issue or to decline toadopt the Ninth Circuit’s test as a whole.’’Maj. Op. at 218.

13. I recognize that the majority does notadopt the Ninth Circuit test verbatim. How-ever, the majority nevertheless errs in treatingits test as an affirmative defense because thatdefense, even as modified, analyzes nothingmore than core issues of likelihood of confu-sion, issues that are within the purview ofplaintiffs’ case.


Holmes held that ‘‘[w]hen the mark is usedin a way that does not deceive the publicwe see no such sanctity in the word as toprevent its being used to tell the truth. Itis not taboo.’’ Id. at 368, 44 S.Ct. 350.Prestonettes is clearly a nominative usecase, and the Supreme Court did not treatthe use as one that had to be justified bythe defendant.

The majority seems to dismiss Presto-nettes on the basis that it is not a newcase. See Maj. Op. at 223 n.4. I am notaware of any rule permitting us to disre-gard Supreme Court precedent simply be-cause it is not as fresh as the Sundaypaper. The majority fails to persuade thatPrestonettes is anything other than a para-digm nominative use case. Indeed, eventhe Ninth Circuit, when formulating itsnominative use test in New Kids, cited,quoted, and relied on Prestonettes whendefining the nature of nominative use:

[W]e may generalize a class of caseswhere the use of the trademark does notattempt to capitalize on consumer confu-sion or to appropriate the cachet of oneproduct for a different one. Such nomi-native use of a mark—where the onlyword reasonably available to describe aparticular thing is pressed into service—lies outside the strictures of trademarklaw: Because it does not implicate thesource-identification function that is thepurpose of trademark, it does not consti-tute unfair competition; such use is fairbecause it does not imply sponsorship orendorsement by the trademark holder.‘‘When the mark is used in a way thatdoes not deceive the public we see nosuch sanctity in the word as to preventits being used to tell the truth.’’ Presto-nettes, Inc. v. Coty, 264 U.S. 359, 368, 44

S.Ct. 350, 351, 68 L.Ed. 731 (1924)(Holmes, J.).

971 F.2d at 307–08 (emphasis in origi-nal).

In any case, we need not turn to theNinth Circuit, nor rely solely on olderprecedent, to find more recent bindingnominative use cases. This Court faced anominative use dispute in G.D. Searle &Co. v. Hudson Pharmaceutical Corp., 715F.2d 837 (3d Cir.1983). In that case, thedefendant used the plaintiff’s METAMU-CIL mark to truthfully state that its prod-uct was ‘‘Equivalent to METAMUCIL.’’Id. at 838. The District Court enjoinedthe defendant from making conspicuoususe of the mark but otherwise permittedtruthful use of the mark. On appeal, weaffirmed that defendant could truthfullyuse the mark. Relying on Prestonettes,we concluded that ‘‘whether one is entitledto refer to a competitor’s trademark de-pends not on where the reference appears,but on whether the reference is truthful.’’Id. at 843 (discussing and quoting Presto-nettes, 264 U.S. at 368, 44 S.Ct. 350). Wefurther held that such use was permissible‘‘so long as it does not contain misrepre-sentations or create a reasonable likeli-hood that purchasers will be confused asto the source, identity, or sponsorship ofthe advertiser’s product.’’ Id. at 841(quotes omitted, emphasis added).14

Three things are obvious in Searle.First, we had no doubt that Prestonetteswas the controlling case for the disputeduse, a use that modern parlance deems‘‘nominative use.’’ See id. at 843 & n. 13(quoting Prestonettes, 264 U.S. at 368, 44S.Ct. 350). Second, we treated the ques-tion before us solely as whether the usewas likely to confuse. ‘‘ ‘[C]ollateral andtruthful references to the trademark of

14. In addition, we affirmed the limited in-junction, which enjoined defendant from set-ting the mark apart in larger or differently

colored type, and required it to use ‘‘b’’ anduse the disclaimer ‘‘a product of G.D. Searle,not a Hudson product.’’ Id. at 839, 843.


another are permissible as long as the‘‘unauthorized’’ reference does not causeconfusion as to the source of the productadvertised.’ ’’ Id. at 841 n. 9 (quotingDelaware and Hudson Canal Co. v. Clark,80 U.S. (13 Wall) 311, 327, 20 L.Ed. 581(1872)) (emphasis added). Third, notablymissing from Searle was any suggestionthat the defendant bore any burden ofsatisfying an affirmative defense. Instead,the question was whether the conductwould ‘‘create a reasonable likelihood thatpurchasers will be confused as to thesource, identity, or sponsorship of the ad-vertiser’s product.’’ Id. at 841 (emphasisadded, internal quotes omitted).15

Like Prestonettes, Searle is a paradigmnominative use case. The mere fact thatthese decisions did not use the term ‘‘nom-inative use’’ when they were decided doesnot make them any less binding. Bothcases make it clear that nominative use isnothing more than a likelihood of confusioninquiry, and neither places upon a defen-dant the burden of negating confusion orof putting forth an affirmative defense.16

II.

This section examines the inner work-ings of the majority’s bifurcated test.Subsection II.A shows that the majority’sapproach permits plaintiffs to establishlikely confusion on as little as on Lapp

factor, and improperly shifts to defendantthe burden of negating confusion. Subsec-tion II.B shows that the majority’s ap-proach is judicially unmanageable becauseit requires courts to consider identical like-lihood of confusion factors on both sides ofthe analysis. Subsection II.C shows thatthe majority’s approach stems from a mis-taken conflation of the statutory defense ofdescriptive fair use, which asks whether adefendant has used a mark in its primarydescriptive sense, with nominative use,which simply asks if a defendant has creat-ed a likelihood of confusion.

A.

The majority’s approach whittles downthis Court’s ten-factor Lapp test beyondrecognition, permitting plaintiffs to showlikely confusion on as little as one factor.The effect of this is to shift to defendantthe burden of negating confusion. I willstart by outlining the majority’s proposedorder of analysis. Although the majorityconcedes that eight of the ten Lapp factorscan be relevant in nominative use cases, itlimits extended discussion to these fourfactors:

1. the price of the goods and otherfactors indicative of the care andattention expected of consumerswhen making a purchase;

15. Notably, in Pebble Beach, where the FifthCircuit declined to adopt the Ninth Circuit’stest verbatim, the Court cited Searle as sup-port for the proposition that ‘‘[t]his right touse a mark to identify the markholder’s prod-ucts—a nominative use—however, is limitedin that the use cannot be one that creates alikelihood of confusion as to source, sponsor-ship, affiliation, or approval.’’ 155 F.3d at546 (citing, inter alia, Searle, 715 F.2d at 842)(emphasis added). Unlike the majority, theFifth Circuit clearly understood Searle to be anominative use case and that the issue waswhether the use created a likelihood of confu-sion.

16. The majority strangely complains that nei-ther Prestonettes nor Searle contain any ‘‘ref-erence to ‘fair use,’ let alone nominative fairuse.’ ’’ Maj. Op. at 223 n.4. Of course theydon’t, because both cases treat the analysis asone of likely confusion, not as an affirmativedefense. And to state the obvious, neithercase was decided before the Ninth Circuitcoined the phrase ‘‘nominative use.’’ Themajority’s approach has the impermissible ef-fect of attempting to overrule Searle, whichcan only be done by the en banc Court.


2. the length of time the defendant hasused the mark without evidence ofactual confusion;

3. the intent of the defendant in adopt-ing the mark; 17 and

4. the evidence of actual confusion.

Maj. Op. at 225–26.18

If likely confusion is shown by plain-tiffs—which as discussed below, might beshown on one Lapp factor alone—then themajority would require defendant to satis-fy its ‘‘nominative fair use’’ defense. Asmildly reworded from the Ninth Circuit’stest, the majority would require defendantto show:

1. Is the use of plaintiff’s mark neces-sary to describe (1) plaintiff’s prod-uct or service and (2) defendant’sproduct or service?

2. Is only so much of the plaintiff’smark used as is necessary to de-scribe plaintiff’s products or ser-vices?

3. Does the defendant’s conduct or lan-guage reflect the true and accuraterelationship between plaintiff anddefendant’s products or services?

Maj. Op. at 228.

I will first consider the majority’s bare-bones Lapp test. Regarding the first fac-tor—the price of the goods and other fac-tors indicative of the care and attention

17. Because nominative use defendants use aplaintiff’s mark rather than adopt it, the moreaccurate phrasing would be ‘‘the intent of thedefendant in using the mark.’’ I have fol-lowed this approach in Section III, infra.

18. To be clear, the panel is unanimous inholding that in an appropriate case, all butthe first two Lapp factors may be relevant tothe nominative use analysis. See Maj. Op. at225–26. Moreover, although the majority fo-cuses on only four of the factors, id. at 225,231–32, it at no point suggests that thosefactors constitute the sole set of factors fornominative use analysis; instead, it notes thatthe analysis will ‘‘include’’ those four factors.See Maj. Op. at 225. Though the majorityexpresses doubt regarding the weight of thelast four Lapp factors to this particular dis-pute, ibid., at no point does it hold that theDistrict Court is prohibited from consideringthem. As it states: ‘‘We find that all of theother Lapp factors, while perhaps not appro-priate for analysis in this particular case,could be analyzed in future nominative usecases, depending on the factual situation.’’Ibid. (emphasis added). Apparently conclud-ing that the parties are actual or close com-petitors, the majority appears to suggest thatthe last four factors will be of little relevancein this case. Perhaps, but as the majorityimplicitly recognizes, the determination of thedegree of dissimilarity and its weight is forthe District Court to determine in the firstinstance.

Section III, infra, outlines how the eightrelevant Lapp factors, including the final four,should be analyzed in a nominative use case,including this case. If the District Court findsthat the parties’ services, customers, advertis-ing, and markets are highly similar or identi-cal—i.e., the more the parties are competi-tors—such findings have no bearing under thelast four Lapp factors in establishing a likeli-hood of confusion. Competitors often makenominative use for purposes of ‘‘comparison,criticism, [or a] point of reference.’’ NewKids, 971 F.2d at 306; see also Searle, 715F.2d at 841 (competitor selling laxatives de-scribed as ‘‘Equivalent to METAMUCIL’’). Innominative use cases, focusing on the parties’status as competitors would lead to false find-ings that all nominative uses are confusing;accordingly, the parties’ possible status ascompetitors has little to no bearing on estab-lishing a likelihood of confusion, and otherfactors must be addressed to analyze likeli-hood of confusion.

But when the parties are not competitors,the last four Lapp factors may become rele-vant in cutting against a likelihood of confu-sion in the nominative use context. The morethe parties’ services, customers, advertising,or markets diverge, the more they reduce thelikelihood of confusion. This is because dif-ferently targeted consumers may see themarks in different contexts, through differentadvertising, and in relation to different goodsor services. These differences may counselagainst a likelihood of confusion.


expected of consumers when making apurchase—the District Court had little tosay, noting that the issue was ‘‘a littledifficult’’ and concluding that real estatepurchases are the most important pur-chases at the time and perhaps in people’slives. Although the District Court’s opin-ion is not clear, it appears that the Courteither found that the factor was neutral orpointed somewhat towards defendant.

The second and fourth factors—lookingto the absence or presence of actual confu-sion—do not assist plaintiffs in this casebecause they stipulated that they wouldnot rely on evidence of actual confusion forthese proceedings. This is unsurprisingbecause actual confusion evidence is inmany cases unavailable and is not requiredto find likely confusion.19

This leaves only one factor, intent. Inthis regard, the majority holds that ‘‘if thecourt finds that the defendant made use ofthe plaintiff’s mark with the very purposeof causing consumers to think the plaintiffendorses or sponsors plaintiff’s good orservice, then the likelihood that consumerswill be confused as to endorsement/affilia-tion is greater.’’ Maj. Op. at 226. Accord-ingly, holds the majority, ‘‘the key inquiryis whether the mark is being used so as toconvey a connection between the partiesthat may not exist.’’ Ibid. I agree withthe majority that the intent inquiry is im-portant in nominative use cases. Indeed,if the plaintiffs in this appeal are to showlikely confusion at all, they will have toprove it through this factor. And consid-ering the parties’ serious dispute over theveracity and context of the defendant’s use

of the plaintiffs’ marks, I have no doubtthat on remand that the intent factor willbe the locus of contention.

But the majority is unclear as to whatfactors the District Court might look inproving intent. It correctly notes thatmere ‘‘[u]se of the mark alone is not suffi-ciently probative of such intent.’’ Maj. Op.at 227. This makes sense because themere existence of nominative use, withoutmore, lacks probative value as to whetherthe use was made with the intent to con-fuse. But the majority provides littleguidance on what might serve as affirma-tive proof of intent in a nominative usecase, noting unhelpfully that the DistrictCourt should consider whether defendant‘‘used the mark with the intent to confusethe public as to the relationship betweenthe defendant and the plaintiff.’’ Ibid.(emphasis in original). But that admoni-tion merely restates the original inquiry,i.e., what evidence proves bad intent?

The majority’s terse discussion of theintent factor begs clarification of the kindof proof that would show intent to confusein a nominative use case. Fortunately,such clarification is not difficult to find inthe majority’s opinion, although its locationis troubling. One might look to the pur-pose of a defendant’s use, which would beprobative of whether the use was made ingood faith or instead with the intent toconfuse. One might look to the promi-nence of the use; if defendant used themark in too-large or stylized form, that

19. Although proof of actual confusion is notrequired, the more such evidence there is, thegreater the likelihood of confusion. VersaProducts Co., Inc. v. Bifold Co. (Mfg.) Ltd., 50F.3d 189, 205 (3d Cir.1995). Conversely, alack of actual confusion evidence does notraise an inference of nonconfusion ‘‘when theparticular circumstances [do not] indicate

such evidence should have been available.’’Id. (alteration in original, quote marks delet-ed). But ‘‘[i]f a defendant’s product has beensold for an appreciable period of time withoutevidence of actual confusion, one can inferthat continued marketing will not lead to con-sumer confusion in the future.’’ Id.


may show an intent to use more than isnecessary, and again show an intent toconfuse. One might also look to the truth-fulness of the use: if the use was truthful,that would suggest a lack of intent toconfuse; conversely, misleading or un-truthful uses of plaintiffs’ marks may beprobative of an intent to confuse.

If the above characterization sounds fa-miliar, it should. Each and every one ofthe majority’s ‘‘nominative fair use’’ prongsis nothing more than an inquiry into thelikelihood of confusion, specifically, wheth-er the use is with the intent to confuse dueto a presence or lack of good-faith purpose,prominence, and truthfulness. The majori-ty’s first ‘‘nominative fair use’’ prong looksto purpose and asks: ‘‘Is the use of plain-tiff’s mark necessary to describe (1) plain-tiff’s product or service and (2) defendant’sproduct or service?’’ This prong asks whydefendant is using the mark, i.e., its pur-pose in using it. By the majority’s formu-lation, if it is necessary for defendant touse the mark to describe both the plain-tiffs’ and its own services, then such use

may be fair.20 But the reason for any‘‘fairness’’ would be because the use wasmade with good intent, a paradigm likeli-hood of confusion inquiry, and one thatfalls within the normal Lapp analysis.

The second inquiry suffers the same in-firmity. It asks, ‘‘Is only so much of theplaintiff’s mark used as is necessary todescribe plaintiff’s products or services?’’This inquiry looks to the prominence ofthe use, asking if the defendant ‘‘took toomuch,’’ i.e., whether it used bigger, boldertype or a logo when plain type might havesufficed. And why is this relevant? Theanswer is clear: the willful overuse ofplaintiffs’ marks may evidence bad-faithintent and is more likely to confuse. In-deed, Searle itself affirmed an injunctionthat enjoined the defendant from settingthe mark apart in larger or differentlycolored type, and required it to use ‘‘b’’and a disclaimer. See 715 F.2d at 839,843. As such, the majority’s prominenceinquiry again is nothing more than part ofthe Lapp analysis.21

20. I further object to the first prong of the‘‘nominative fair use’’ defense because of itsconjunctive requirement that a plaintiff’smark be used to describe both plaintiff anddefendant. Although ‘‘nominative’’ use re-quires that the mark be used at a minimum torefer to plaintiff, proper nominative use mayor may not also use a plaintiff’s mark to referto the defendant’s goods or services. Nomi-native use may occur ‘‘where the defendantuses a trademark to describe the plaintiff’sproduct, rather than its own.’’ New Kids, 971F.2d at 308. Sometimes the use of the markmay be, as noted in New Kids, merely for‘‘comparison, criticism, [or a] point of refer-ence.’’ Id. at 306. In any case, I would notethat even under the majority’s phrasing, ‘‘nec-essary’’ under the first prong does not mean‘‘indispensable’’—‘‘the court need not findthat the use of the mark is indispensable inorder to find this factor fulfilled.’’ Maj. Op.at 229.

Another troubling problem that may arisefrom the first prong’s conjunctive phrasing isthat the majority may have unintentionally

foreordained the outcome of these proceed-ings through its passing statement that ‘‘[de-fendant] is not attempting to use [plaintiffs’]marks to refer to [defendant’s] own services,but rather is using the marks to refer to[plaintiffs’] services.’’ Maj. Op. at 225. Themajority appears to suggest that defendantdid not use the marks to refer to itself. I willemphasize that this determination is not forthis Court to make and should be made by theDistrict Court upon remand.

21. The majority is also unclear regarding thephrasing of the second prong, which asks‘‘[i]s only so much of the plaintiff’s mark usedas is necessary to describe plaintiff’s productsor services?’’ Maj. Op. at 228 (emphasis add-ed). Later in the opinion, the majority states‘‘the proper focus under this prong is onwhether only so much of plaintiff’s mark as isreasonably necessary to identify plaintiff’sproduct or service has been used by defen-dant.’’ Id. at 230 (emphasis on ‘‘reasonably’’added). There may be worlds of differencebetween what is ‘‘necessary’’ and what is


Like the first two prongs, the majority’sthird prong is again nothing more than alikelihood of confusion inquiry. It asks,‘‘Does the defendant’s conduct or languagereflect the true and accurate relationshipbetween plaintiff and defendant’s productsor services?’’ By its own terms, the thirdprong looks to the truthfulness of the useof the mark and the defendant’s conduct indescribing the relationship between theparties and their goods/services. But thistruthfulness inquiry is identical to the stat-utory test of infringement under LanhamAct § 32(1)(a), which asks whether the useis ‘‘likely to cause confusion, or to causemistake, or to deceive.’’ Lanham Act§ 32(1)(a), 15 U.S.C. § 1114(1)(a) (empha-sis added). Similarly, Section 43(a)(1)(A)asks whether the use is ‘‘likely to causeconfusion, or to cause mistake, or to de-ceive as to the affiliation, connection, orassociation of such person with anotherperson.’’ Id. § 43(a)(1)(A), 15 U.S.C.§ 1125(a)(1)(A) (emphasis added). By itsplain language, the Lanham Act looks tothe truthfulness of the parties’ affiliationas crucial parts of the infringement analy-sis. In other words, truthfulness is a partof the likelihood of confusion analysis, andthe majority’s formulation to the contraryis irreconcilable with the statute.

Moreover, the majority’s third fair useprong is substantively identical to its own‘‘intent’’ Lapp factor. The majority’s thirdfair use prong asks whether ‘‘the defen-dant’s conduct or language reflect[s] thetrue and accurate relationship betweenplaintiff and defendant’s products or ser-vices.’’ Maj. Op. at 228 (emphasis added).This inquiry is identical to the majority’sintent test under Lapp, which states ‘‘thekey inquiry is whether the mark is beingused so as to convey a connection betweenthe parties that may not exist.’’ Id. at 226(emphasis added). Whether one calls usesthe words ‘‘connection’’ or ‘‘relationship,’’ itis obvious that the majority’s third fair useprong and its Lapp intent inquiry are iden-tical.22 As the majority places identicallikelihood of confusion factors on bothsides of the equation, there can be nodoubt that the majority expects defendantto prove nonconfusion.

In sum, the ‘‘nominative fair use’’ de-fense is nothing more than a venture intoLapp territory, namely, intent as shownthrough the defendant’s purpose, promi-nence, and truthfulness in connection withplaintiffs’ marks.23 Although it is trou-bling that the majority would permit plain-tiffs to establish likely confusion based onone Lapp factor alone, it is even more

‘‘reasonably necessary.’’ Although I read themajority’s second statement as clarification ofthe meaning of ‘‘necessary,’’ the majority isnot clear on this important point.

22. The identity of the Lapp analysis and theaffirmative defense is further underscored bythe majority’s additional descriptions of theLapp ‘‘intent’’ inquiry: whether ‘‘the defen-dant made use of the plaintiff’s mark with thevery purpose of causing consumers to thinkthe plaintiff endorses or sponsors plaintiff’sgood or service,’’ Maj. Op. at 226 (emphasisadded), and whether defendant ‘‘used themark with the intent to confuse the public asto the relationship between the defendant andthe plaintiff.’’ Id. at 227 (emphasis in origi-nal).

23. Just like the majority, the Ninth Circuit’sfirst two ‘‘nominative fair use’’ prongs looksto purpose and prominence. The Ninth Cir-cuit’s third prong embraces truthfulness andmore, unabashedly requiring that ‘‘the usermust do nothing that would, in conjunctionwith the mark, suggest sponsorship or en-dorsement by the trademark holder.’’ NewKids, 971 F.2d at 308. Try as it might todress up the Ninth Circuit’s test in newclothes, the test was and remains an inquiryinto likelihood of confusion, and the majorityrepeats the Ninth Circuit’s error of placingthe burden of negating confusion onto thedefendant.


disturbing that the majority would requirethe defendant to dispel that findingthrough an affirmative defense looking tothe same likelihood of confusion consider-ations.24

B.

Not only does the majority’s test runafoul of KP Permanent and the LanhamAct by placing upon defendant a burden ofnegating confusion, but it is also judiciallyunmanageable. On remand, the DistrictCourt will be required to consider thepurpose, prominence, and truthfulness ofthe defendant’s conduct as part of theplaintiffs’ likelihood of confusion case. Itwill also be required to look to the samefactors when entertaining the defendant’saffirmative defense. This makes no sense.The majority points to no meaningful dis-tinctions between the kinds of ‘‘intent’’ thatmight be relevant to likelihood of confu-sion, and those which might be relevant tothe affirmative defense. This is becausethey are identical. I respectfully submitthat District Courts will not be able to

make heads or tails of how to conduct theanalysis.

The practical effect of the majority’sbifurcated analysis is that the nominativefair use defense will almost always beavailable only in cases where it is unneed-ed. Where a court finds the use to bemade with the intent to confuse, then afinding of likelihood of confusion will oftenfollow. According to the majority, in sucha case, a defendant could then assert nomi-native fair use. However, such a defensewould be hollow indeed, because the entire‘‘nominative fair use’’ defense asks wheth-er the use was made with the intent toconfuse. Thus, a defendant found to havecaused likely confusion due to illegitimatepurpose, overly prominent use, or untruth-fulness will be unable to assert a successful‘‘nominative fair use’’ defense. Converse-ly, in cases where the alleged nominativeuse is made with good-faith purpose, is notoverly prominent, and is truthful, then aplaintiff will likely fail in proving its caseof likely confusion. In that case, the de-fense will be unneeded.25

24. Ironically, the majority criticizes the Dis-trict Court in Liquid Glass Enterprises, Inc. v.Dr. Ing. h.c.F. Porsche AG, 8 F.Supp.2d 398(D.N.J.1998), suggesting that the Court ‘‘con-flated the [New Kids ] test with the plaintiff’sburden of proving likelihood of confusion.’’Maj. Op. at 218. Yet the majority’s teststands guilty of the same accusation that itaims at Liquid Glass.

25. Trying to find some use for its affirmativedefense, the majority argues that ‘‘[i]t is thecircumstance in which a court does not findbad intent but does find confusion that anominative fair use defense will be most use-ful.’’ Maj. Op. at 227 n.7 (some emphasisadded). Ironically, such a use would be al-most the only circumstance where a nomi-native fair use defense would even theoretical-ly be available. If likelihood of confusionexists due to bad intent, the defense will beunavailable because the intent is bad. If like-lihood of confusion doesn’t exist, then thedefense won’t needed. This leaves the cir-cumstance noted by the majority: if likeli-

hood of confusion is somehow shown despitea lack of bad intent, defendant can point to itsgood intent to prevail. But considering that akey inquiry in a nominative use case is whythe defendant has used the plaintiff’s mark(purpose) and how it has been used (promi-nence and truthfulness), it is hard to imaginea case where necessary, non-prominent, andtruthful use would be likely to confuse.

Such a finding under the facts of this casewould appear to require highly unsophisticat-ed consumers and an abundance of actualconfusion evidence. But this hypotheticalpossibility—notably, circumstances not pres-ent in this case, possibly rendering the majori-ty’s approach as nothing more than dicta—does not warrant the majority’s invention ofan affirmative defense that duplicates likeli-hood of confusion intent inquiries and placesupon defendant an affirmative burden ofshowing good faith. In any case, the majori-ty’s unlikely scenario of likely confusion de-spite good faith may not warrant a completebar on such use, and under appropriate cir-


The result will certainly be that thenominative fair use defense, as formulatedby the majority, will remain unavailable inmany of the cases where it is needed dueto a bad-faith intent to confuse. The Su-preme Court frowned heavily on this ab-surdity in KP Permanent where it held,‘‘ ‘[I]t defies logic to argue that a defensemay not be asserted in the only situationwhere it even becomes relevant.’ ’’ 125S.Ct. at 549 (quoting Shakespeare Co. v.Silstar Corp., 110 F.3d 234, 243 (4th Cir.1997)). The Court further noted that ‘‘itwould make no sense to give the defendanta defense of showing affirmatively that theplaintiff cannot succeed in proving someelement (like confusion).’’ Id. Yet this isexactly what the majority has done.

C.

The majority’s bifurcated analysis ap-pears to arise from a failure to recognizethat KP Permanent addresses several is-sues, only the first of which is applicable tothis case, namely, the Supreme Court’sholding that a defendant bears no burdenof negating confusion. KP Permanent’ ssecond holding concerns matters peculiarto the statutory descriptive fair use test, atest with no bearing in the present context.

The first issue in KP Permanent, andthe one that controls the outcome of this

case, regards which party bears the bur-den of demonstrating a likelihood of con-fusion. As the majority recognizes, KPPermanent clearly holds that this burdenbelongs to a plaintiff alone. Nothing inthe Lanham Act requires a defendant tonegate a likelihood of confusion.26 As heldby the Supreme Court, the Act requiresthat the burden remain on plaintiffs: inthat case, the Court held the statute‘‘places a burden of proving likelihood ofconfusion (that is, infringement) on theparty charging infringement even whenrelying on an incontestable registration.’’KP Permanent, 125 S.Ct. at 548. Thus,held the Court, ‘‘the burden of provinglikelihood of confusion rests with theplaintiff.’’ Id. at 550.

The Court further held that it would beincoherent to place upon defendant theburden of showing nonconfusion:

[A] look at the typical course of litiga-tion in an infringement action points upthe incoherence of placing a burden toshow nonconfusion on a defendant. If aplaintiff succeeds in making out a primafacie case of trademark infringement,including the element of likelihood ofconsumer confusion, the defendant mayoffer rebutting evidence to undercut theforce of the plaintiff’s evidence on this(or any) element, or raise an affirmative

cumstances, may merit at most a constrainedremedy that continues to permit such truthfuluse, such as ordering clarifying language or adisclaimer, as happened in Searle. Cf. 715F.2d at 839, 843.

26. Lanham Act § 32(1)(a) states that ‘‘[a]nyperson who shall, without the consent of theregistrant—(a) use in commerce any repro-duction, counterfeit, copy, or colorable imita-tion of a registered mark in connection withthe sale, offering for sale, distribution, or ad-vertising of any goods or services on or inconnection with which such use is likely tocause confusion, or to cause mistake, or todeceive TTT shall be liable in a civil action by

the registrantTTTT’’ Lanham Act § 32(1)(a),15 U.S.C. § 1114(1)(a). Similarly, Section43(a)(1)(A) asks whether the use is ‘‘likely tocause confusion, or to cause mistake, or todeceive as to the affiliation, connection, orassociation of such person with another per-son.’’ Id. § 43(a)(1)(A), 15 U.S.C.§ 1125(a)(1)(A). Compare this language tothe majority’s affirmative ‘‘nominative fairuse’’ defense—which asks whether ‘‘defen-dant’s conduct or language reflect[s] the trueand accurate relationship between plaintiffand defendant’s products or services.’’ Maj.Op. at 228 (emphasis added). It’s the sameinquiry, and it improperly places the burdenof negating confusion on defendant.


defense to bar relief even if the primafacie case is sound, or do both. But itwould make no sense to give the defen-dant a defense of showing affirmativelythat the plaintiff cannot succeed in prov-ing some element (like confusion); allthe defendant needs to do is to leave thefactfinder unpersuaded that the plaintiffhas carried its own burden on that point.A defendant has no need of a court’strue belief when agnosticism will do.

Id. at 549. For the reasons notedabove, I conclude that the majority’s bifur-cated analysis and the phrasing of its‘‘nominative fair use’’ defense violate thisfirst and clear holding of KP Permanent,by placing the burden of showing noncon-fusion on defendant.

A second issue in KP Permanent, incontrast, has no bearing to this case. Thatissue was what role, if any, a finding oflikely confusion may play when a defen-dant asserts the statutory affirmative de-fense of descriptive fair use under LanhamAct § 33(b)(4). The statutory defenseasks whether ‘‘the use of the name, term,or device charged to be an infringement isa use, otherwise than as a mark, TTT of aterm or device which is descriptive of andused fairly and in good faith only to de-scribe the goods or services of such party,or their geographic origin.’’ Lanham Act§ 33(b)(4), 15 U.S.C. § 1115(b)(4). Unlikethe present dispute, which asks whetherdefendant created a likelihood of confusionby using the plaintiffs’ marks to refer atleast in part to plaintiffs, a descriptive fair-use defense would ask whether a defen-dant may use a trademark (e.g., ‘‘BESTPRICE’’) in its primary descriptive sense(e.g., ‘‘we offer the best price on breakfastcereal’’). In the latter case, a thresholdinquiry and key distinguishing fact is that

defendant would allege that it is not usingthe plaintiff’s marks to refer to plaintiff atall, but is instead using the mark in adescriptive way without reference to theplaintiff.

In KP Permanent, the alleged infringerargued that it used variations of the term‘‘micro color’’ to refer to its goods in adescriptive way without reference to thetrademark owner. The question beforethe Court was whether a finding of likelyconfusion would foreclose assertion of thestatutory descriptive fair use defense.27

The Court held that it would not, conclud-ing that ‘‘our holding that fair use canoccur along with some degree of confusiondoes not foreclose the relevance of theextent of any likely consumer confusion inassessing whether a defendant’s use is ob-jectively fair.’’ 125 S.Ct. at 550.

Despite the majority’s suggestions tothe contrary, the Court’s second holding isnot earth-shattering—the Court merelyheld that even though a finding of likelyconfusion would not foreclose assertion ofthe statutory descriptive fair use defense,courts might nonetheless factor the extentof confusion into determining whether thepurportedly descriptive use was not fair.In the context of descriptive fair use, theCourt’s statement makes sense becausethe extent of confusion might be relevantto determining whether a the mark was‘‘used fairly’’ or ‘‘otherwise than as amark.’’ Lanham Act § 33(b)(4), 15 U.S.C.§ 1115(b)(4). But descriptive fair use andnominative use are very different animals:in descriptive fair use, the key inquiry iswhether a plaintiff’s mark is being used ina descriptive sense, but in nominative usecases, courts ask one question alone: is the

27. Underlying the case was a long-standingcircuit split over whether a finding of likelyconfusion foreclosed a statutory descriptive

fair use defense. See KP Permanent, 125 S.Ct.at 547 (collecting cases).


defendant’s use likely to confuse? 28 In-deed, even while noting that some caseshad treated the extent of confusion as arelevant consideration in descriptive fairuse cases, the Supreme Court immediatelyand unequivocally cautioned that ‘‘[c]er-tainly one cannot get out of [such cases]any defense burden to negate [confusion]entirely.’’ 125 S.Ct. at 549.

It makes no sense to force the squarepeg of KP Permanent’ s second holdingabout descriptive fair use into the roundhole of nominative use. Nominative use isnothing more than an inquiry into likelyconfusion. By reading KP Permanent topermit a bifurcated analysis where confu-sion is the central focus on each side, themajority reads KP Permanent to hold thatboth: 1) defendant bears no burden ofnegating confusion, and 2) defendant bearsa burden of negating confusion. Normight the majority justify its defense bycharacterizing it as an inquiry into goodfaith. As stated by the Supreme Court,‘‘Nor would it make sense to provide anaffirmative defense of no confusion plusgood faith, when merely rebutting theplaintiff’s case on confusion would entitlethe defendant to judgment, good faith ornot.’’ Id. Indeed, good faith is merely theflip side of bad faith, and as discussedpreviously, the intent factor has alwaysremained part of the plaintiff’s case underLapp.

The majority’s attempt to rely on KPPermanent in support of its bifurcated

analysis is further and finally belied by thefact that the Court took pains to note thatit was not addressing nominative use. Seeid. at 546 n. 3 (‘‘Nor do we address theCourt of Appeals’s discussion of ‘nomi-native fair use.’ ’’). The majority’s trans-plantation of the Supreme Court’s com-ments about descriptive fair use into thenominative use context is no more effectivethan replacing a healthy human heart withone from another species. In short, classicfair use is a defense. Nominative use isnot.

III.

Merely criticizing the majority’s ap-proach would be hollow commentary with-out providing in addition an outline on hownominative use analysis can proceed in afashion consistent with the Lanham Act,KP Permanent, Prestonettes, Searle, andthis Court’s long-standing Lapp test.Needless to say, because the nominativeuse analysis is an inquiry into the likeli-hood of confusion, the burden must remainwith plaintiffs, although—as in any trade-mark infringement case—defendant re-mains free to put on arguments and evi-dence to rebut those put forth by plaintiffs.I start as I must with reference to ourstandard ten factors, as phrased in ourdecision in A & H Sportswear, Inc. v.Victoria’s Secret Stores, Inc.:

(1) the degree of similarity between theowner’s mark and the alleged infringingmark;

28. Descriptive fair use and nominative use‘‘are quite different from each other and bearlittle resemblance except that both are typesof non-infringing use.’’ 2 J. Thomas McCar-thy, McCarthy on Trademarks and UnfairCompetition, § 11:45 (4th ed.2005). As statedby the Ninth Circuit in distinguishing nomi-native use from descriptive (or classic) fairuse:

The distinction between classic and nomi-native fair use is important for two reasons:

(1) classic and nominative fair use are gov-erned by different analyses; and (2) theclassic fair use analysis only complementsthe likelihood of customer confusion analy-sis set forth in Sleekcraft, whereas the nomi-native fair use analysis replaces the Sleek-craft analysis.

Cairns, 292 F.3d at 1150. In other words,descriptive fair use looks beyond confusionconsiderations whereas nominative use looksonly to a likelihood of confusion.


(2) the strength of the owner’s mark;

(3) the price of the goods and otherfactors indicative of the care and atten-tion expected of consumers when mak-ing a purchase;

(4) the length of time the defendant hasused the mark without evidence of actu-al confusion arising;

(5) the intent of the defendant in adopt-ing the mark;

(6) the evidence of actual confusion;

(7) whether the goods, competing or notcompeting, are marketed through thesame channels of trade and advertisedthrough the same media;

(8) the extent to which the targets ofthe parties’ sales efforts are the same;

(9) the relationship of the goods in theminds of consumers, whether because ofthe near-identity of the products, thesimilarity of function, or other factors;

(10) other facts suggesting that the con-suming public might expect the priorowner to manufacture both products, orexpect the prior owner to manufacture aproduct in the defendant’s market, orexpect that the prior owner is likely toexpand into the defendant’s market.

237 F.3d 198, 215 (3d Cir.2000) (discuss-ing Interpace Corp. v. Lapp, Inc., 721 F.2d460 (3d Cir.1983)).29 The majority is cor-rect that the factors may be tailored to thecircumstances: ‘‘[n]ot all factors will berelevant in all cases,’’ and ‘‘different fac-tors may properly be accorded differentweights depending on the particular factu-al setting.’’ Id.

I agree with the majority that Lappfactors one and two are unlikely to carryany weight in a nominative use case. In acase where a defendant is using a plain-tiff’s mark to refer to the plaintiff, the usewill always be identical.30 Moreover, itmakes little sense to give weight to factortwo, the strength of the mark, as nomi-native uses will be made with both weakand strong marks.

The majority focuses in this case on onlyfour of the ten Lapp factors (factors threethrough six), but agrees that in somecases, the remaining factors (seventhrough ten) might be relevant. I believeit is important to explore circumstanceswhere Lapp factors seven through ten maybe pertinent. In this case, much of theparties’ services, customers, advertising,and markets are similar and may be insome regards identical. Although the de-gree of identity or similarity is for theDistrict Court to determine in the firstinstance, the majority correctly concludesthat to the extent the parties share identi-cal services, customers, advertising, andmarkets, such identity lacks value in deter-mining the likelihood of confusion. SeeMaj. Op. at 225 (‘‘it would almost be ex-pected that [defendant] and most otherdefendants in a nominative use case wouldmarket themselves through a media inwhich the marks to which they are refer-ring would be easily recognized and havemeaning or relevance, namely channelssimilar to those used by plaintiff’’). How-ever, to the extent that the parties’ ser-vices, customers, advertising, and marketsare different, factors seven through tenmay counsel against a likelihood of confu-

29. This updated formulation post-dates Lappand addresses both competitive and non-com-petitive uses.

30. ‘‘When a defendant uses a trademark nom-inally, the trademark will be identical to theplaintiff’s mark, at least in terms of the wordsin question. Thus, application of the [normal

multi-factor] test, which focuses on the simi-larity of the mark used by the plaintiff and thedefendant, would lead to the incorrect conclu-sion that virtually all nominative uses areconfusing.’’ Playboy Enterprises, Inc. v.Welles, 279 F.3d 796, 801 (9th Cir.2002).


sion. Compare this case to one involving anewscaster. What if the hypotheticalnewscaster used plaintiffs’ logos as part ofa newscast about plaintiffs? Regardless ofFirst Amendment issues, the differencesbetween the parties’ services would them-selves counsel against a likelihood of con-fusion. Thus, as recognized by the majori-ty, factors seven through ten are not perse irrelevant, and the door remains open tothem in relevant cases.

In light of these considerations, I recom-mend that in cases involving colorablenominative use—i.e., where a defendanthas used a plaintiff’s mark to refer at leastin part to the plaintiff—the District Courtconsider Lapp factors three through ten,slightly reworded and renumbered onethrough eight:

1. The price of the goods or servicesand other factors indicative of the careand attention expected of consumers whenmaking a purchase. As noted by the ma-jority, this remains a relevant inquiry.The cheaper the goods or the less sophisti-cated the consumers, the more likely thata use may confuse. This factor will re-quire careful attention by the DistrictCourt on remand. People buying and sell-ing homes might be likely to take especialcare in choosing an agent, but the use heremay also give rise to concerns under the‘‘initial interest confusion’’ doctrine. SeeCheckpoint Sys., Inc. v. Check Point Soft-ware Techs., Inc., 269 F.3d 270, 292 (3dCir.2001) (holding initial interest confusionactionable under the Lanham Act); seealso Lamparello v. Falwell, 420 F.3d 309,316 (4th Cir.2005) (quoting Checkpoint andnoting limitations on initial interest doc-trine where use was for criticism or com-mentary rather than profit).

2. The length of time the defendant hasused the mark without evidence of actualconfusion arising. ‘‘If a defendant’s prod-uct has been sold for an appreciable periodof time without evidence of actual confu-sion, one can infer that continued market-ing will not lead to consumer confusion inthe future.’’ Versa Prods., 50 F.3d at 205.‘‘The longer the challenged product hasbeen in use, the stronger this inferencewill be.’’ Id.

3. The intent of the defendant in usingthe mark. As noted by the majority, ‘‘thekey inquiry is whether the mark is beingused so as to convey a connection betweenthe parties that may not exist.’’ Maj. Op.at 226. Although I disagree with the ma-jority’s bifurcation and duplication of theintent inquiry, there can be no doubt thatthis inquiry is crucial and may include, asnoted by the majority, ‘‘myriad’’ factors.See Maj. Op. at 227. Thus, the DistrictCourt should inquire as to the nature ofthe defendant’s conduct and the accuracyof its use of the plaintiffs’ marks (akin tothe majority’s third ‘‘nominative fair use’’prong). These inquiries extend further tolooking to the purpose of the use (akin tothe majority’s first ‘‘nominative fair use’’prong),31 and to whether defendant ‘‘tooktoo much’’ with an eye towards creatingconsumer confusion (i.e., the second ‘‘nomi-native fair use’’ prong), such as a competi-tor using a plaintiff’s mark in logo form,larger type, or different colors, in a waydesigned to draw attention above and be-yond proper nominative use. Indeed, thisfactor was expressly considered in Searle,where we enjoined the defendant from set-ting the mark apart in larger or differentlycolored type, and required it to use ‘‘b’’and a disclaimer. 715 F.2d at 839, 843.

31. But as noted, I cannot ascribe to the ma-jority’s conjunctive phrasing that would re-quire that the use be necessary to describe

both a plaintiff’s product or service and adefendant’s product or service. See supranote 11.


4. The evidence of actual confusion.The ‘‘more evidence of actual confusionthat a plaintiff can muster, the strongerthe likelihood of confusion in the future,but lack of evidence of actual confusion,’’ atleast where the time period is short or thecircumstances do not indicate that suchevidence should have been available, ‘‘doesnot raise an inference that there is nolikelihood of confusion.’’ Versa Prods., 50F.3d at 205. Here, plaintiffs concede forpurposes of this motion that there is nosuch evidence, which for present purposesweighs in defendant’s favor.

5. Whether the goods or services, com-peting or not competing, are marketedthrough the same channels of trade andadvertised through the same media.Where the parties are competitors and thechannels of trade or media are identical,this factor has no relevance because com-petitors by nature use identical channelsand media. However, to the extent thatthe parties are not competitors, as in thecase of the hypothetical broadcaster, orthe parties use differing channels of tradeor media, this factor lessens the likelihoodof confusion because of differences in thetarget audiences and the places where con-sumers will see the marks used. In thiscase, defendant and plaintiffs are not iden-tical competitors but appear to compete insome respects. They also appear to usewebsites geared towards similar audiences.The extent that the parties are or are notcompetitors, or do or do not have identicalchannels of trade or media, should be de-termined by the District Court upon re-mand. If identical, then this factor has noweight. If only slightly different, the fac-tor has some but little bearing in reducingthe likelihood of confusion. And the great-er the differences, the more weight theDistrict Court should place on this factor.

6. The extent to which the targets ofthe parties’ sales efforts are the same.

The rationale behind factor six is similar tothat for factor five. In a nominative usecase, the mere fact that the parties targetidentical customers has no bearing. Butto the extent the targeted customers dif-fer, that lessens the likelihood of confusion.Upon remand, the District Court shoulddetermine the extent to which the targetedcustomers are identical or not.

7. The relationship of the goods or ser-vices in the minds of consumers, whetherbecause of the near-identity of the prod-ucts or services, the similarity of function,or other factors. The same rationale ap-plies. The mere fact that the goods orservices are identical has no probative val-ue because competitors may oftentimesmake otherwise nonconfusing nominativeuse. But to the extent the goods or ser-vices differ, that lessens the likelihood ofconfusion. Upon remand, the DistrictCourt should determine the extent towhich the parties’ services are similar ornot.

8. Other facts suggesting that the con-suming public might expect the prior own-er to provide both products or services, orexpect the prior owner to provide a prod-uct or service in the defendant’s market,or expect that the prior owner is likely toexpand into the defendant’s market. Thisfactor asks whether consumers would ex-pect plaintiffs to offer defendant’s services,or offer services in defendant’s market, orto expand into defendant’s market. Thesame rationale applies here as for factorsfive, six, and seven; namely, to the extentthat the consuming public expects plain-tiffs to offer defendant’s services, to offerservices in the defendant’s market, or toexpand into the defendant’s market, thisfactor has no bearing. But the less theconsuming public would expect such con-duct from plaintiffs, the more this factorcuts against a likelihood of confusion.


Upon remand, the District Court shouldconsider this factor as well.

IV.

The test outlined above best satisfies KPPermanent’ s requirement that plaintiffsbear the burden of establishing a likeli-hood of confusion. It also satisfies therequirements of the Lanham Act, Presto-nettes, and Searle, which provide that nom-inative use is use that is not likely toconfuse. Unlike the majority’s approach,this test is judicially manageable because itdoes not bifurcate and duplicate the analy-sis.

The majority objects that it would beunfair to deprive a defendant of a nomi-native fair use defense. This complaintrings hollow: defining ‘‘fair use’’ in termsof nonconfusion does not give the majoritylicense to place on defendant’s shoulders aburden that it should not bear. It istherefore the majority’s formulation that is‘‘unfair.’’

Moreover, the majority is simply wrongthat defendants will have no way of under-cutting a charge of infringement. As theSupreme Court noted in KP Permanent inresponse to this very objection, a defen-dant remains free to ‘‘offer rebutting evi-dence to undercut the force of the plain-tiff’s evidence on’’ likelihood of confusion.KP Permanent, 125 S.Ct. at 549. As theSupreme Court held, ‘‘all the defendantneeds to do is to leave the factfinder un-persuaded that the plaintiff has carried itsown burden on that point. A defendanthas no need of a court’s true belief whenagnosticism will do.’’ Id.

V.

For the reasons discussed above, I can-not join in the majority’s formulation of abifurcated test because it runs afoul of theLanham Act, KP Permanent, Prestonettes,

and Searle, each of which make it clearthat the burden in nominative use casesremains with plaintiffs. Because the ma-jority’s test impermissibly places the bur-den of negating confusion on defendant,and is in addition judicially unmanageable,I respectfully dissent. I nevertheless joinin the judgment remanding the case forfurther proceedings.

,

Brian Keith MOORE, Petitioner–Appellant,

v.

Philip PARKER, Warden,Respondent–Appellee.

No. 03–6105.

United States Court of Appeals,Sixth Circuit.

Argued: March 10, 2005.

Decided and Filed: Oct. 4, 2005.

Background: Following affirmance ofstate convictions for first-degree robbery,kidnapping and capital murder, and sen-tence of death, state prison inmate soughtfederal habeas relief. The United StatesDistrict Court for the Western District ofKentucky, Thomas B. Russell, J., deniedpetition, and inmate appealed.

Holdings: The Court of Appeals, Cook,Circuit Judge, held that:

(1) no unreasonable application of federallaw occurred when state appellatecourt found no prejudice in trial attor-ney’s deficient failure to impeach wit-ness;

(2) defense attorneys did not perform defi-ciently by failing to seek out additional

CENTURY 21 REAL ESTATE CORP. v. LENDINGTREE, INC. 211 · CENTURY 21 REAL ESTATE CORP. v....

Documents

Transcript of CENTURY 21 REAL ESTATE CORP. v. LENDINGTREE, INC. 211 · CENTURY 21 REAL ESTATE CORP. v....