Service Chain Management – Optimizing Service...

http://fitzsimmons.CRMproject.com 1

White Paper

WRITTEN BY

James A. Fitzsimmons is the

William H. Seay Centennial

Professor of Business

at the University of Texas.

Tom Meredith serves as senior

vice president and chief

financial officer for

Dell Computer Corporation.

Douglas Morrice is an

Associate Professor in the

Management Science and

Information Systems

Department at the University

of Texas at Austin.

G. Edward Powell is the founder

and chairman of Pointserve,

a company that specializes in

mobile workforce optimization.

William H. Rice is a

board member of Pointserve,

chairman of @Outcome (an

Internet healthcare company)

and a venture partner with

Murphree Ventures.

Devanath Tirupati is an

Associate Professor of

Management at the University

of Texas at Austin.

James Fitzsimmons, Douglas Morrice and Devanath Tirupati

The University of Texas at AustinEdward Powell and William Rice

PointServe

Envision1.0Discover2.0Personalize3.0Interact4.0Connect5.0

Service Chain Management –Optimizing Service Delivery

Supply Chain Management is a holistic approach to delivering manufactured

products to the end customer. Using advanced information technology and

optimization software, the elements of the supply chain are coordinated to

achieve efficient distribution unavailable in traditional logistics systems.

Service Chain Management is an analogous systems approach designed for

manufacturing and delivering a service; it addresses the challenges and

opportunities for organizations facing the prospect of establishing a service

chain or improving an existing chain.

IntroductionOver the past decade, the topic of Supply ChainManagement has received much attention.

1The

supply chain paradigm, enabled by information technology, has eliminated non-value-adding activities and established interdependent relation-ships among trading partners.

2In this article, we

propose the study of supply chains for services andcall it Service Chain Management; service chains canbenefit from many of the same principles developedfor goods supply chains with at least similar, if notgreater, gains in productivity.

Service sector growth and supply-side competi-tion underscore the importance of Service ChainManagement. For example, the home health careindustry faces unprecedented growth in demand overthe next 10 years: suppliers of home health care services must now do more with less. Under deregulation, market competition is forcing manyutilities to streamline operations; globalization confronts, these organizations with potential markets orders of magnitude larger than their existing customer base.

3Companies unwilling to

compete on cost alone must redesign their existingsupply chains to deliver the service portion of theirproduct offerings in a cost-effective manner.

4

Some of the challenges facing Service ChainManagement organizations include the opportunity forbi-direction optimization, the role of conversion anddistribution, effective use of productive capacity, theperishable nature of the service product, the level ofcustomization, the sequence of transactions impliedby the service offering, and the level of uncertainty;the degree to which they impact Service ChainManagement is often extreme: for many services, production and consumption must be simultaneous.

We illustrate many of the concepts here usingone of the most important segments of the servicesector: services delivered through a mobile workforce. This service segment of the economy iscomprised of 15 million workers who produce a trillion dollars of the United States GDP (15% of theGDP).5 Figure 1.0 provides a comparison of a goodssupply chain with a service supply chain for a mobileworkforce that captures many of the issues that wewill discuss in this paper.

Defining the Service ChainIn defining the service chain, we take the view of the external customer. The processes involved indelivering a service to a customer define the servicechain. Customers generally consider the request and

Defying the Limits: Reaching New Heights in Customer Relationship Management2

White PaperConnect5.0Interact4.0Personalize3.0Discover2.0Envision1.0

fulfillment portions of a service as a singlesystem. See Figure 1.0.)

Many companies and systems can create a ”virtual” service system. Most ser-vice companies separate service-requestprocesses from service-production processes,with inherent inefficiencies.6 One strategyof Service Chain Management is to integratethese two separate systems into one.

In healthcare, there is a virtual systemthat connects the various service providers(doctors, hospitals, pharmacies, etc.) inresponding to patient needs; it may be amulti-step process involving an HMO, a primary care physician, a specialist, a hospital-based-out-patient radiology depart-ment and a national laboratory. Treatment issimilar; it may include independent pharmacies, medical equipment suppliers,home healthcare providers and return visitsto the physician’s office to measureresponses to treatment. Even in ”integrateddelivery networks” there are often surprisingly separate units that rarely function efficiently from the patient’s view.

Bi-Directional Optimization Bi-directional optimization implies doingwhat is best from both the customer and the service enterprise perspective. Inservice production, customers often ”co-produce” the service. Direct customerinvolvement facilitates bi-directional optimization, a simultaneous optimizationof both supply and demand for the service;this provides individualized service for thecustomer and cost-effective service for theservice enterprise.

Knowledge management is a strategy toorganize and use available information,experience and expertise to create a compet-itive advantage for a business. Technologyinnovations now enable a near instantaneousinclusion of customer requirements and preferences into the knowledge managementstrategy of service companies.

Every customer has unique expecta-tions, wants, needs, preferences and desires.These factors are the basis for buy decisions,the perception of quality, the chance forrepeat business and the value of branding.

FIGURE 1.0 Comparison of a Goods Supply Chain and a Service Supply Chain

Examples of Service ChainManagement in Home HealthcareHome healthcare visits by 2001 will exceed600 million visits provided by over 300,000health workers.10 This large mobile work-force is comprised of workers with a varietyof skill sets and knowledge: registerednurses, licensed vocational nurses, nursingaids, physical and occupational therapists,respiratory therapists, etc.

Patients receiving home healthcare areeven more varied, with different needs andpreferences. A bedridden patient may prefer a same-gender nurse for bathing procedures and other very personal carematters. Some patients do not speak fluentEnglish and require language-appropriateproviders. Because of the personal natureof healthcare, most patients prefer toreceive their care from a limited number offamiliar providers.

Given the nature of the business ofhome healthcare, the solution to the ”trav-eling salesman problem” for this market isexceedingly complex; traditional approachesof organizing and providing home healthcareare significantly sub-optimized.

With this background, tools and tactics that can automate and optimizecost-drivers are becoming an essential partof the survival strategy of home healthcareenterprises.

The mission-critical functions of homehealthcare agencies are generally organizedby a series of paper-based systems. Newpatients are received via telephone referrals. Nurses go to the patient’s homeand perform assessments that address thecomplete medical and social environmentof the patient. This drives the creation of acare plan which defines the ”visit plan,” acalendar-driven work plan which maps outwhen the patient is seen, what happensduring each visit, what skill set is associatedwith each visit and when the patient’s status requires re-assessment.

On a weekly basis, supervisory nursesat the home healthcare agency office takethese care plans and write down all the”jobs” for each patient by day, for the current week, to create a master schedule.


The dynamic availability of individualcustomer requirements and preferencesadds a new dimension to the knowledgemanagement possibilities of service companies. This capability is transforminghow a mobile workforce provides highlypersonalized services at significantly lowercost than before.

For mobile service, a ”time-window” isreserved for each customer to be availablefor a service worker to perform the request-ed service. Customers see generous time-windows as a major inconvenienceequivalent to wasting their time for thebenefit of an inefficient service provider.Currently, mobile workforce serviceproviders ask customers to be available forextended time windows because routes,customer sequencing, and interval serviceprovider job status updates are not builtinto the information system. Service ChainManagement uses forecasting data to con-struct an initial daily plan for each worker.It offers several service appointment choicesto customers based on a pre-optimized calculation of an optimized solution for thewhole group of service workers in a city orany defined geographic area. This opti-mization process can reduce customer timewindows from 5 hours to 1 hour, improveresponsiveness to individual customer preferences, eliminate paper-based docu-mentation and billing forms, and decreaseoperational costs from 5% to 15%.

The Role of Conversion and DistributionConversion is the process of creating valuefor the customer. For goods, conversion isprincipally located in the factory; for services, including mobile services, it is typically located at a customer site. In manufacturing, production rarely involvesthe customer actively; in services, the cus-tomer often is a co-producer. In the typicalmanufacturing supply chain, conversionoccurs early in the process, in the servicechain it occurs late in the process. A typi-cal factory will focus on a limited number ofstandard products, a mobile service workergenerally has considerable discretion andwill be expected to customize the service tothe customer’s needs. These conversionprocess differences distinguish between theproduction of goods and services and haveimplications for the strategies used for operational optimization.

Distribution of most manufacturingvalue is associated with strategies to haveproducts available on the retail store shelfwhen a customer reaches makes a purchase.Distribution logistics and delivery of goodsare a relatively low-cost component of mostmanufactured items. Sales channels thatare long-term, relatively static partnershipscreate additional opportunities to decreasethe relative cost of a distribution network.

In mobile service production, thenature and character of the service is

Providers generally continue working withpatients they have already seen. This mas-ter schedule is then reformed into a dailytask list for each provider. There is often afrenzy of activity just prior to schedulerelease associated with ”holes” in theschedule and supervisors ”fishing” for a will-ing provider to take just one more patient.

Once the weekly schedules are finalized and printed, providers pick uptheir schedules and the week begins. Homehealthcare providers work autonomously.The time of home visits is generally notconsidered when producing the schedules.Providers are left to make arrangementswith patients based on mutual agreementand providers often intermingle personaltasks with patient visits. This schedulefreedom, while attractive, builds invisibleinefficiency into the system.

Once homecare visits are completed,visit documentation is returned to theagency office where the billing process canbe initiated.

Service Chain Management of this system consists of automation and optimization of the important cost driversassociated with the mission critical function of home healthcare. It begins withthe automation of the care planningprocess using a PC-based visit tool that isused as the patient intake tool (this is thesame as a graphical user interface for a callcenter). Individual patient address, clinicalrequirements and personal preferences areentered and can now be actively and consistently incorporated into the enter-prise knowledge management system.

On the same PC is the database thatorganizes provider information by skills,preferences, licensing, current training andcertification, salary or contract paymentdata, provider home address, as well as fax,pager and mobile phone numbers. Also onthe same PC is an advanced software modulethat can process the information for enterprise operational optimization. In amatter of minutes it explores the require-ments, preferences, and location of everypatient and the skills, training, licensure,

FIGURE 2.0

120 Days 30 Days 72 Hours 10 Minutes10 Seconds

Supply Chain(Traditional)

DirectManufacturing

Model

ServiceChain End-User

Time from final customization to delivery to an end-user



usually determined relatively close to thetime of delivery. Every moment spent indistribution equals a lost moment of pro-ductive capacity. This relationship reversalgives a significantly enhanced importanceand value to the role of distribution in themobile service sector. Traditional inventorystocking and demand anticipation strategiesof most of the manufacturing sector areimpossible for the mobile service sector. The productive capacity of a mobileservice worker cannot be placed into inventory for distribution at a later time.

Managing Productive CapacityFor mobile workers, a primary considerationof productive capacity is the amount of timespent between jobs. Because value is creat-ed predominantly through time at the customer site, time spent traveling is lostproductive capacity. Service ChainManagement can decrease the travel time forthe enterprise significantly, which greatlyincreases workforce productivity throughproviding more time for workers to be withcustomers. Strategies to improve productivecapacity of the service worker also includetransfer, replacement and embellishment.

• Transfer makes knowledge available tocustomers so that value can be transferred

with very low cost. One example is a Web-based FAQ database. This information-transfer tool can replace more expensivenursing resources and is always available.Patients can access disease-specific areasof a Web site and obtain informationabout side effects of medicines, explanations of symptoms, or proceduresto follow.

• Replacement substitutes expensiveresources with cheaper ones. For patientsneeding blood pressure measurementsthree times daily, an automated measure-ment system might be substituted for anurse visit

• Embellishment of customer skills toenable self-service is a third strategy toenhance the productive capacity of thehome healthcare system. Teaching apatient or family member to change a surgical dressing is appropriate in certaincircumstances. This embellishment of customer skills requires an incrementalamount of additional nursing time initial-ly, but significantly decreases the use ofnursing resources if the wound is chronicand requires daily changes for an extendedperiod of time. In a typical scenario, anurse might check the healing wound

preferences and location of every provider.It can run millions of scenarios that matchpatients and providers in different combinations and sequences. The complexmathematical modeling, advanced artificialintelligence algorithms and extensive solution-space exploration are invisible tothe user.

Cost and quality drivers are an essen-tial component of the software; the systemhas ”dials” which can be adjusted for each.For example, continuity of care is an impor-tant factor in patient satisfaction. Havingthe same provider for a given patient 95%of the time versus 80% of the time mayhave a very significant impact on the costof providing services for the entire enterprise. Similarly, each preference of thepatient and provider can be given a relativeweight and the effect of these decisions ondifferent solutions can be compared. Thissystem-wide understanding of the impactof certain policy decisions on the actualcost of providing services is a new possibil-ity for the home healthcare enterprises.

It is our view that Web-based connec-tivity will shortly replace or significantlychange the myriad of care documentationand disparate billing systems which are thefocus of most of the information technologyinvestment in the home healthcare industry.

Service Chain Management for a homehealth agency can anticipate servicedemand using a database that containspatient age, diagnosis, co-diagnosis, carerequirements, individual preferences andpatient location. This accessible data candrive a whole range of productive capacityissues for the agency; a predominance ofdiabetes among Spanish-speaking patientsin a metropolitan area may drive hiring andtraining of local providers with a certainskill-set, leading to shorter travel-times,better provider continuity, improvedresponsiveness and lower overall enterprisecosts. Forecasting demand with greaterdetail in the data can significantly impactthe way an enterprise builds and maintainsit workforce to meet the individual needsand preferences of the patient population

40

35

30

25

20

15

10

5

0

Monthly Total Non-Billable Visits

AfterService Chain Management

BeforeService Chain Management Software

Nov Dec Jan Feb

FIGURE 3.0 Monthly Total Non-Billable Visits


every three days instead of daily, resultingin a 66% decrease in the amount of nursingtime required to care for the patient.

These three approaches to improving theproductive capacity of a service systemmaximize the efficiency of systems such as”capitated” healthcare services. Untilrecently, reimbursement in home health-care was based on the number of visits provided. The financial incentive in themarketplace was not to enhance productivecapacity, but to maximize the number ofpatient visits. The result is that homehealthcare has been, generally, as expensive as the system would allow, shortof fraudulent, unnecessary visits. TheMedicare payment system is currently transitioning the home healthcare paymentsystem to a ”prospective payment system.”Under this new payment system, thepatient’s diagnosis drives a single, ”flatfee” payment for the entire disease processepisode. Extending the productive capacityof nurses has now become an importantstrategy of the business of home healthcare.7

Management of Perishability Unlike the manufacturing sector, it is impos-sible to capture service worker time to storefor future service demands. The productive

capacity of a service worker is limited to thetime he or she is at the customer site, withthe right tools and skills, with knowledge of the customer’s requirements and preferences. Management of perishability isthe approach used in Service ChainManagement to minimize the negativeimpact of idle time on the productive capac-ity of the distributed service workforce.

For the mobile workforce, managingperishability has two foci. The first is a timeallocation system that offers time windowsto customers based on ”best use” of workers.In Service Chain Management systems thatuse dynamic schedule optimization software,schedules are constructed and revised upuntil the very last possible moment. Servicesystems with no communication infrastruc-ture allowing ”real-time” dispatch producedaily schedules at the beginning of theworkday. Service systems with mobile datacommunications coupled with advancedoptimization algorithms may produce ”real-time” schedules so that a worker onlylearns of the next job at the completion ofthe current job. The dynamic allocation ofjobs among workers assures that idle (perishable) time is minimized.

Managing perishability also involvesthe process of training, refining andextending the skills and capabilities ofworkers. Potential idle time of workers can,

it serves. Forecasting is a knowledge man-agement tool which is used to convert historical care experience, geographicalpatient and provider data, and cumulatedpreference and requirement data to system-atically build a workforce that is the mostappropriate and cost-effective for thepatient population served by a homehealthcare enterprise.

Similarly, service chain bi-directionaloptimization of the patient-agency require-ments, capabilities, and preferences is asystem that incorporates new levels ofresponsiveness into the provider system.This is based on the process of automationand optimization of the closely coupledconversion and distribution aspects of theservice process, a distinctly different coupling than is existent in the supplychain for manufactured goods.

Service Chain Management is criticalin the home healthcare industry; transfer-ring, replacing and embellishing productivecapacity is an important response to thenew capitated payment system. Providertraining, skill set enhancement, and cross-functionality should be part of theservice process. Given that much of this ismodular, self-paced, ”distance learning”activity, incorporating these activities intothe enterprise plan and cost model is anopportunity to create an ”inventory hedging” model for home healthcare.

Service Chain Management softwareincludes modules for forecasting demand,automating the patient-intake and careplanning processes, and a global optimiza-tion module for deploying the provider work-force based on the agency-cost-model,patient requirements, location and preferences, as well as provider skills andlocation. The automation and optimizationtools can produce a 6% reduction in overalloperating costs while improving patient satisfaction, worker satisfaction and agencyresponsiveness to patient preferences.

In one agency, total weekly schedul-ing time was reduced from 144 hours to 36hours. Missed visits, (approximately 2.5%historically, usually due to paper-based

6000.0

5000.0

4000.0

3000.0

2000.0

1000.0

Total Weekly Miles Driven

Without Service ChainOptimization

With Service ChainOptimization

FIGURE 4.0 Total Weekly Miles Driven

customer selects from a variety of productsbuilt one to three months, ago). The bestdirect model supply chains offer dramaticimprovement by permitting customization towithin 72 hours of the moment of consump-tion. In services, the service chain can bedesigned to permit some level of customiza-tion right up until the moment of consump-tion. As Fisher poses, customization at thislevel requires a supply chain that is flexibleand responsive to changes in demand.Flexibility and responsiveness can comefrom managing capacity, having more high-ly-skilled, ”cross-trained” workers, and/orbi-directional optimization.9

Managing a Sequence of TransactionsWhile most goods require a single salestransaction, many services require asequence of transactions over a period oftime. Such arrangements are common inhome health care and equipment mainte-nance. This presents many challenges andopportunities. It requires more sophisticat-ed forecasting techniques (e.g., life cyclemanagement strategies) for predicting thetiming and the nature of a series of relatedevents, and also requires a database containing the history of each transactionsequence, i.e., a case history. When thistype of information is coupled with realtime information and bi-directional optimization, a service chain can be flexibleand responsive to changing customer needsthroughout the sequence of transactions.



by design, be directed to these activities bythe process of work schedule optimizationin Service Chain Management. When activelyincorporated into the business process ofthe service worker, significant service pro-duction capacity time can be ”prospectivelyreclaimed.” This service-inventory-hedgingstrategy benefits the enterprise by directingthese skill-enhancing activities into idletime slots and creates potential availabilityof the worker during the time when tradi-tional worker training efforts would haveconsumed productive capacity. The mobileworker with mobile data communicationtools can consume training, certificationand testing material online, maximizing thishedging strategy.

Perishability can also be managed byadjusting capacity. Capacity adjustmentstend to take longer and be more costly thanadjusting inventory levels to meet changesin demand. Consequently, in a service chain,adjustment lags resulting from changes indemand tend to be longer. As a result, thepotential for the bullwhip effect is greater.8

CustomizationLike goods, services range from commoditiesto highly customizable products. However,the extent to which a service is customizableis greater. Figure 2.0 compares the time-from-final- customization to delivery to anend-user of a good and a service. For thetraditional make-to-stock supply chain, theopportunity for customization ends any-where from 30 to 120 days out (i.e., the

transcription errors) were completely eliminated. Non-billable visits (patient notat home) could be greatly reduced due toinclusion of patient-availability parameterswithin the global optimization database(Figure 3.0) .

The total weekly miles driven wasreduced by 25% (Figure 4.0). Decreasing travel time for the mobile workforce signifi-cantly expands productive capacity of theenterprise; the optimization module of the Service Chain Management software produces this while still meeting all patientand provider requirements. It also includes anagency-managed balance of preferences andcost metrics for the patient and the enterprise. Service Chain Management toolsproduce bi-directional optimization of thescheduling process, and increase productivecapacity by decreasing travel miles (time) andtransferring replacement and embellishmentof provider skills as a planned approach to thebusiness of home healthcare. Perishabilitymanagement in the home healthcare sectorwill become a more important strategy inworkforce optimization as the mobile datainfrastructure of this industry grows.


In order to mitigate uncertainty for asequence of service transactions, contractsare often used. However, the terms of thesecontracts require not only legal expertisebut also reliable forecast information onboth demand and supply of the service inorder to establish pricing arrangementsthat are profitable for the seller and affordable for the buyer.

Management of UncertaintyUncertainty results from having to predict asequence of service transactions, and fromthe operating environment; many servicesare produced in uncontrolled operatingenvironments, and there exists a greaterlevel of uncertainty in service delivery. Forexample, in the oilfield services industry,mobile workers conduct operations in anoutdoor environment around the world, facing uncertainty from weather, terrain,wildlife and property permitting.

Although the level of uncertainty forservice delivery may be high, it must beassessed to develop proper hedging strate-gies against risk. Forecasting and simulationtools can be combined with the best information available on local conditions(e.g., terrain, weather, etc.) to assess riskand then develop strategies to manage it.

ConclusionService Chain Management is a set of toolsthat are used to model, automate andoptimize the delivery of services.Although conceptually similar to supplychain management, the ”means and measures” of Service Chain Managementare different because the nature of servicedelivery and goods production is fundamentally different. Highlightingthese differences are the methods used tofacilitate bi-directional optimization, andthe management of productive capacityand perishability.

It is probable that these ideas will beimportant in other service sectors of theeconomy. Bundling of services with goodsand services will increasingly be a differen-tiation strategy for producers of both goodsand services. It is likely that, in the future,more and more of these bundled serviceswill be directed toward the fast-growingaging population. Effectively serving thispopulation will increasingly include services provided by mobile workers.

Footnotes1 Edward Miller, ”Engineering/manufacturing: The

Importance of Supply Chain Management,” Inform, vol.

13, no.4, 1999.

2 Li Dong and Christopher O’Brien, ”Integrated Decision

Modeling of Supply Chain Efficiency,” International Journal

of Production Economics, vol.59, no. 3, 1999, pp.147-157. 3 James Truesdell, ”Deregulation of Utilities is a Two-Edged

Sword,” Supply House Times, vol. 41, no. 7, 1998, pp.

57-58.4 Chris Storey and Christopher Easingwood, ”The

Augmented Service Offering: A Conceptualization and

Study of Its Impact on New Service Success,” Journal of

Product Innovation Management, vol. 15, no. 4, 1998,

pp. 335-351.5 James Franklin, ”Industry Output and Employment

Projections to 2006,” Monthly Labor Review, vol. 120,

no. 11, 1997, pp. 39-57.6 Markku Tinnila and Ari Vepsalainen, ”A Model for

Strategic Repositioning of Service Processes,”

International Journal of Service Industry Management,

vol. 6, no. 4, 1995, pp. 57-80.7 John Ferman, ”The New Medicare Outpatient PPS,”

Healthcare Executive, vol.14, no. 1, 1999, pp. 48-49.8 Hau Lee, V. Padmanabhan, and Seungjin Whang, ”The

Bullwhip Effect in Supply Chains,” Sloan Management

Review, vol. 38, no. 3, 1997, pp. 93-102.9 Marshall L. Fisher, ”What is the Right Supply Chain for

Your Product?,” Harvard Business Review, vol. 75, no. 2,

March-April 1997, pp. 105-116.10 Mary Wagner, ”Multi-Unit Providers Survey: Providers of

Home Healthcare Expand and Consolidate to Meet Rising

Demand,” Modern Healthcare, vol. 18, no. 23, 1998, pp.

80-84.

Service Chain Management – Optimizing Service...

Documents

Transcript of Service Chain Management – Optimizing Service...