MANAGING RISKSIN ENTERPRISE SYSTEMSIMPLEMENTATIONS

JUDY E. SCOTT AND IRIS VESSEY

Two companies wereat risk. One survivedwhile the other failedafter installing anenterprise system(ES),SAPR/3.Atthetime of its R/3 imple-mentation, begin-

ning in 1995, Dow Corning Incoqjoratedwas a $2.5 billion producer of siliconeproducts. The company was facing com-petitive pressures as well as lawsuits worth$2 billion due to well-publicized problems with sili-cone breast implants. Existing systems were frag-mented and focused on specific departments, makingit difficult to present a common face to the cus-tomcr.Thc company decided that its survivaldepended on reengineering its business processes tobecome a truly global company, an objective itbelieved could be met only witb appropriate informa-tion systems. It created Business Processes and Infor-mation Technology (BPIT) to support reengineeringby incorporating responsibility for business processesand IT within tbe same organizational unit.

In preparation for its initial implementations—threepilot sites in the U.K.̂ —-the company gave up some ofits best busine.ss practices, ;uid paid litde attention tomanaging change witbin the organiz^ition. It did nottrain the project team appropriately and did not havethe benefit of a consulting partner. Yet the companyand the system survived.

This scenario can be compared witb FoxMeyerDrug Corporations R/3 implementation. FoxMeyersbusiness was wholesale drug distribution, an industryrendered very competitive and relatively unstable byhealth-care reforms ofthe early 1990s. Hoping to gaina competitive advantage, FoxMeyer initiated Its $65million R/3 project in 1993. At the same time,FoxMeyer installed extremely ambitious warehouseautomation software, which was to be interfaced witb

W H A TDETERMINESIF A SYSTEM

IMPLEMENTATIONWILL BE

SUCCESSFUL?

R/3 at a new warehouse. FoxMeyerexpected to save $4{)-$5O million dollarsannually from the project, as well as togrow rapidly and gain market share [5].

FoxMeyers plans did not work out.After its major customer, Phar-Mor,went bankrupt in May 1993,FoxMeyer signed a major new cus-tomer. University Fiealthsystem Con-sortium (UHC). However, the contractrequired major changes to the project.Costs soared to over $100 million; and

in August 1996 FoxMeyer filed fbr Chapter 11 bank-ruptcy protection, after taking a charge of $34 mil-lion the previous month for inventory and ordermix-ups. Following liquidation of its major assets inNovember 1997, FoxMeyers trustee sued AndersenConsulting, SAP, and Deloitte for $500 million eachin July and August 1998. The case against Deloittewas dismissed in May 1999.

There are many reports of software project failures.In addition to widely publicized figures on system fail-ures in general. The Standish Ciroup International esti-mates that 90% of R/3 projects run late [12]. Hencethese are substantive problems that keep occurring inthe industry and that need to be addressed.

Both the Dow Corning and FoxMeyer implementa-tions involve the same softwaie; they took place at sim-ilar times (1994-1996) and therefore with similarpackage capabilities and ES knowledge; and both com-panies were suffering significant industr)' pressures.The risks in ES implementations are highlighted here(see the sidebar "Characteristics of ES and ES Proj-ects"), and we seek to answer the questions: Why didDow Coming's implementation succeed whileFoxMeyers did not?; Which risks are critical?; Are theretrade-of̂ s or contingencies among implementationrisks? and. What can be done to mitigate the effects ofthese risks? We do this by presenting a model of riskfactors in ES implementation and illtistrating the rela-

74 April !002/Vol 45 No 4 COMMUNICATIONSOFTHEACM

ExternalBusine55Context

CompetitiveEnvironment

Unstablecompetitiveenvironment

CollaborativeEnvironmentUnstablecollaborativeenvironment

CooperativeEnvironmentUncertaincooperativeenvironment

Organizational Context

tionsliips and trade-ofFs inherent in the model usingthe Dow Corning and FoxMeyer cases. With our ini-tial understanding, we seek to isolate strong effects,such as company failure, tather than nuances that causesimple cost overruns. Information on Dow ComingsR/3 implementation comes from teaching cases [9]prepared from in-depth interviews. All informationreported here on the FoxMeyer implementation isavailable from online sources.

Model of Risk Factors in ESImplementationsBecatise the implementation of a cross-funaional ESresults in major organizational changes, our model ishased on forces influencing change. Three kinds offorces, external and/or internal to the company, result indifferent degrees of change [6], Those forces result in:movement of the organization as a whole; movement ofparts of the organization; and movement of individualsand groups. In our model, because changes due to sys-tems can also be influenced by factors in the businessenvironment, we envisage four levels: external businesscontext, organizational context, information systemscontext, and ES project (see Figure 1).' Figure 1 isdepicted as a series of concentric circles representing thedifferent levels of analysis; Figure 2 presents the inter-relationships inherent in the model.

Moving from the outer to the inner circles reflects a

The t.itcgoni*s within each level were deveioixid by symht-siamj; litiTdiure iin snftwan:project risks, risks in TT pnije< ts. anil critical success (actors in ES implementations. Siotand Vcs.sey present a tumplete ilescripticin of motlel dcvclcipment in [11|.

focus that changes from strategic totactical. Factors that influencechange in the external businesscontext during an ES implementa-tion project may have profoundimplications for the implementa-tion; that is, the effects are likely tobe far-reaching. However, they willoccur quite infrequently, depend-ing on the stability of the particularbusiness environment. While fac-tors associated with the project willbe less devastating than strategicissues, they will occur much morefrequently (Relationship 1 in Fig-ure 2).

We also envisage two further setsof interrelationships based on theinteractions among model levels.What we are proposing here arespecific types of contingency effects.First, we expea that changes athigher, strategic levels will necessi-tate responses at lower, tactical lev-els. For example, it may be possibleto overcome a change in the exter-

nal business context by making changes in projectscope and/or project schedule (Relationship 2).

Second, if certain of the more far-reaching strategicfactors are handled well, then tactical fectors that arehandled less well may not endanger the project.Although the implementation would proceed less effec-tively and efficiently, the ES implementation could stillbe achieved. However, the reverse would not necessar-

Figure l . Model ofrisk factors in ESimplementations.

Rgure 2. Interrelationships in the risk iBCtor modeL

COMMUNICATIONS OF THE ACM April 2002/Vol 4S. No 4 75

Phase

Prior to initialimplementations

PilotImplementations

Follow-up

Year

1991

1994

1995

1996

1997

1998

Month

Early

June

February

September30

December

Milestone

Global Order Entry System (GOES) ProjectGlobal process consensus dif^cult

Need to restructure around processes rather than functions

Decision to implement R/3 and Project PRIDE (Process Reengineeringthrough Information Delivery Excellence)

1. Established Process and Information Technology Board (PITB)responsible for setting strategic directions for global informationtechnology

2, Combined IT function with business process reengineering functioninto a unit called Business Processes and Infor mat ion Technology (BPIT)

3.Appointed Charlie Lacefield. the then-VP of Manufacturing andEngineering to manage the new organizational unit

Formation of global team of 40 operations and IT specialists fromaround world

Restructure sub-teams from eight to four to match four key operationalprocesses

New project manager, Ralph Reed

Three U.K. pilot implementations on time but cut scope and there werecommuntcation and change management problems

Global Workstation Project—new distributed infrastructure

Formed area teams in Americas. Europe, and AsiaEstablished PRIDE academy in Midland, Ml—global team to train area teamsFormed think tanks to determine future processes, aid in change

management

Full implementation

Table 1. Important milestones in the Dow Corning ES

implementation history [9].

ily be true^—even if the tactical factors were handledwell, inappropriate handling of thestrategic factors could doom theproject (Relationship 3).

A further set of relationships canbe found within the organiz;itionalcontext factors that contribute toimplementation risk. The most dif-ficult factor to address, and there-fore that tor which the risk ishighest, is a change in organiza-tional culture, followed by firmstrategy, organizational structure,business processes, knowledge/skills, and IT infrastructure [1](Relationship 4.)

Analysis of the DowCorning and FoxMeyer R/3ImplementationsWhat does our model tell us aboutthe risks in the Dow Corning andFoxMeyer implementations? Wefirst outline the similarities in thetwo implementations followed bythe differences. We then use ourmodel to explain the differences: why FoxMeyer didnot survive its R/3 implementation and whether Dow

Corning might have survived insimilar circumstances. In this way,we examine the choices the compa-nies made and the trade-offs thatmay have made a difference to theoutcomes. Tables 1 and 2 presentimportant milestones in the historyof the Dow Corning and FoxMeyerimplementations, respectively, whilelable 3 summarizes the risks associ-ated with the implementations.

Similarities in the Dow Com-ing and FoxMeyer R/3 Imple-mentations. From the viewpoint ofthe external business environment,both Dow Corning and FoxMeyerwere operating in "threatened"environments leading up to theirES purchase decisions. Dow Corn-ing was facing costly lawsuits andincreased competitive pressures,while FoxMeyer was facing decreas-ing margins and high transactionvolumes. From the viewpoint of

org-.inizational context, both companies had well-defined business strategies and viewed IT as key to sup-porting them. Because both companies viewed

Phase

Pre-implementation

Implementation

Aftermath

Year

1993

1994

1995

1996

1997

1998

1999

Month

MayJuly

September

January

July

January toFebruary

January toApril

May

August

FebruaryJulyAugustOctoberMarchFallJulyAugust

May

Milestone

Lost major customer, Phar-Mor Inc.Planning begins on Delta ProjectTesting scalabi ity on Digital and HP servers

Contract with SAP

Contract with Andersen ConsultingHP chosen as hardware platformImplementation of Delta Project begins

Contract with UHC Systems

UHC warehouses openinterface problems result in millions of dollars in inventory losses

Implementation proceedsFl and SD live for UHC

Washington Courthouse warehouse scheduled to go live

VVashington Courthouse warehouse live;damaged inventory and losses from duplicate shipments cost $15.5 million

COO Thomas Anderson asked to resign$34 million charge for inventory and order mix-upsChapter 11McKesson agrees to pay $80 million for FoxMeyerChapter 7Asset liquidationSues Andersen ConsultingSues SAPSues DeloitieCase against Debi t te dismissed

Table 2. Important milestones in the FoxMeyer's

ES implementation history [2I.

7 6 April 2002/Vol 45 No A COMMUNICATIONSOFTHEACM

integrated business processes as the key to their success,both decided to implement an ES. Both companiesrestructured their organizations prior to or in the sametimeframe as initiating their R/3 implementations.

From the viewpoint of implementing an ES, both

Table 3. Summary of risk factors in Dow Coming's and

FoxMeyer's implementations.

Risk Factors

External BusinessContext CharacteristicsCompetitiveEnvironment

CollaborativeEnvironment

CooperativeEnvironmentOrganizationalContext CharacteristicsIT Infrastructure

Knowledge/Skills

BusinessProcesses

OrganiiationalStructure

Firm Strategy

Organizationalculture

Informatlan SystemtChordCterfrttciESChampion/Sponsor

Steering Committee

Package fit

ES Knowledge/Skills

New Technology

Non-ESData

Interfaces

ES ProjectCharacteristicsProject Focus andScope

Pro|ect Management

Organizational ChangeManagement

Dow Corning

Threatened by breast implant lawsuits;competitive pressures

Relatively stable

Relatively stable

Radical change required

Lacked needed skills;training on R/2 inappropriate, inadequate;established PRIDE academy

Originally planned to reengineer;in the end, accepted R/3 processes

Centralized prior to R/3 decision;still needed process-oriented approach

High-growth strategy

Established Tirm;certain changes under way prior to

R/3 decision open culture invitedcommunication;

loyalty

Charlie Lacefield: VP and ExecutiveDirector of Business Processes andInformation Technology (BPIT)

realistic expectations of technology;move forward gradually;small wins

Process and Information TechnologyBoard (PITB)

responsible for setting strategicdirections for global IT

Concerns with processing largenumbers of global users

Lacking;minimal use of consultants

Not used previously for such a largenumber of global users

Omitted bolt-ons. for example, to barscanners, and so forth, to meetproject deadlines

Reduced scope when needed

Strong project leadership

Formed "think tanks" to determinefuture processes, aid in changemanagement

FoxMeyer

Unstable; cut throat competition;acquisitions and mergers

Unstable; lost major customer (Phar-Mor).Signed a replacement customer (UHC)

during implementation. {Did not reactappropriately; see Project Characteristics)

Relatively stable

Radical change required

Lacked needed skills

Originally planned to reengineer;in the end, accepted R/3 processes

Centralized prior to R/3 decision;still needed process-oriented approach

High-growth strategy

Established firm;certain changes under way prior to R/3

decision;less open culture;inadequate communication;lack of loyalty;turnoverproblems

Thomas Anderson: Chief OperatingOfficer;

unrealistic expectations of technology

If in place, did not play a strong leadership

Attempted to assess transaction volume;in the end, proved to be inadequate

Lacking in both FoxMeyer andimplementation partner.

Andersen Consulting

Not used previously for distribution;large volumes of transactions

Conversion problematic;numerous data errors (large $ losses)

Problematic; numerous data errors(large $ losses)

Inappropriate reaction to new customer;compressed schedule; did not testadequately

Who was In charge? Project leadershipinadequate

Had some exposure to change; neededimprovement

companies had a proponent at high levels within theirrespective corporations. There was some doubt aboutthe fit of the R/3 package to both companies' needs:Dow Corning was unsure whether the software wouldultimately support the number of users it anticipatedfor its global operations, while FoxMeyer was uncertainwhether the system could process the required numberof daily transactions. FoxMeyer ran simulations and

determined that RJ5 could, indeed,process the volumes needed; DowCorning did not address the soft-wares limitations in the implemen-tations reported.

Acquiring the necessary ES skillswas problematic for both compa-nies. Dow Coming's employeesreceived R/2 (mainframe) training,instead of the clienc/sei-ver R/3 ver-sion, because R/3 training was notyet available. Similarly, FoxMeyer'straining appears to have beeninadequate.

Both companies needed to con-vert from their existing centndizedmainframe infrastructure to a dis-tributed client/server infrastructure.Both companies implementedvanilla versions of the software,although both implemented lessthan expected due to time pres-sures: FoxMeyer had intended tomake process changes by customiz-ing the system; Dow Corning hadintended to add on other software(for example, to automate autho-rization notices for large purchases).

From the viewpoint of the proj-ect itself, neither company appearedto have very good project planningin place. Further, neither companyappeared to expend a great deal ofeffort on organizational changemanagement. Dow Corningfocused on redticing the length ofthe transition period out of concernfor the project team, whileFoxMeyer had problems with per-sonnel at its warehouses.

Differences m the Dow Com-ing and FoxMeyer R/3 Imple-mentations. Although bothcompanies faced pressures in theirrespective industries, FoxMeyer'ssituation changed during the proj-

COMMUNICATIONSOrTHE ACM April 3002/Vol 45 No 4 7 7

ect while Dow Coming's remained stable. FoxMeyertherefore needed to make adjustments to factors atlower levels in the model to ensure a successful imple-mentation {Relationship 2). The major differencebetween the two companies at the organizational con-text level was in their organizational culture, which,according to our model, has the greatest influence onbringing about organizational change. Dow Corninghad an open culture that invited communication. Infact, Dow Comings value statement: "articulated itscommitment to an open and creative culture that rec-ognized employees as the primary source of ideas,actions, and delivery of performance. Dow Coming'sconsistent growth performance had resulted in a stablework environment that reinforced the loyalty of itsemployees" [9, part A].

FoxMeyer's organizational culture was less open.When FoxMeyer brought forward its Implementationin 90 days as a restilt of its contract with UHC,employees did not voice their concerns to manage-ment. The lack of open communication meantemployees were not particularly loyal to FoxMeyer,which cost the company substantially. For example,project personnel thought they could obtain muchhigher salaries with their newly gained R/3 knowledge;and employees at existing warehouses, knowing theywere going to be out of work, left in droves prior to theimplementation of the automated warehouse.

A further difference between the two implementa-tions lies in the use of consultants. FoxMeyer engagedAndersen Consulting as its implementation partner,

while Dow Corning decided not to use consultantsbecause its culture encouraged employees to try newthings. Dow Coming's employees therefore: "... strug-gled to understand the methodology for simultane-ously learning the software, designing processes, andconfiguring the system. They were particularly chal-lenged by the highly integrated nature of the software,which they cited as Its greatest strength and its greatestweakness.... They spent much time hands-on with theSAP/R3 software exploring what it could and cotildnot do" [9, part Bl.

There were also marked differences in the way thetwo companies viewed technology. FoxMeyer hat!unrealistic expectations of what R/3 could do for it (see[8]): "Insiders say top management was so overopti-mistic about computerization that it recklessly under-bid contracts, expecting electronic efficiencies to lowercosts enough ro make the deal profitable" [2].

Dow Corning, on the other hand, was prepai'ed tomove forward gradually, foregoing its business processreengineering vision to get the system up and runningto reap the benefits of a common, global system priorto seeking more sophisticated processes. The notion ofpursuing "small wins" is regarded as a critical successfactor in process reengineering 17].

At the information systems level, FoxMeyer's projectleadership appears to have been weak. Although theyhad a project champion (COO, Anderson), the projectdid not appear to have an influential steering commit-tee nor a strong project manager. Dow Coming, on theother hand, had a very strong leader (VP and Fxecutive

Characteristics of CS and £S Projects

enterprise systems have been in vogue since the

early ta rnid-1990s when a number of vendors

developed cross-functional integroted systems on

client/server platforms. Many companies seized the

opportunity to implement such systems, which pro-

vided them with both business and technology odvan-

tages. Such systems permit companies to reorganize

their business processes to support the new organiza-

tional paradigm, for example, of lean production, or

just-in-tlme manufacturing. They also support com-

mon, global business processes, and therefore factli-

tate data integration across the enterprise. Further,

the client/server-based IT infrastructure is more flex-

ible and scalable than mainframe systems.

Implementing an ES is o major undertaking. The

size and complexity of the endeavor is influenced by

whether the implementation is focused on the value

chain, and therefore involves sales and distribution,

materials management, and production planning

modules; ar whether it is focused on supporting the

value chain, and therefore involves financials and/or

human resources modules. The size ond complexity of

the implementation is further influenced by the

strategy chosen to roll out the modules (for example.

Big Bong, phased, pilot, or some combination

thereof).

Contrary to traditional wisdom, current recom-

mendations are that companies accept the pracesses

embedded in the package [3]. Because af their cross-

functional nature, ES implementations have signifi-

cont impacts on business processes and therefore an

employees' work lives. Hence, ES implementations

impact ond are impacted by organizational factors

such as the climate for change. U

78 April !002/Vol 45. No 4 COMMUNICATIONSOFTHEACM

The majont}/ of project failiirej(h not J tern from teehiioiogyLkuie^i per je but frommanagement ukiuejthe unplenientatum.

Direaor of BPIT, Lacefield), as well as a planning com-mittee. Further, although neither company managedorganizational change well, Dow Comings culture wasmore conducive to acceptance of change than wasFoxMeyer's.

A further difference between the two implementa-tions is that FoxMeyers difficulties with interfacing sys-tems forced it to take a S34 million charge to coveruncollectible costs telated to shipping and inventoryproblems 12]. There are no reports of such problems atDow Corning.

Model-based Analysis of FoxMeyer's Reaction toEnvironmental Change. By far the most ctucialchange that either company experienced was FoxMey-er's loss of a substantial customer early in its R/3 imple-mentation. Because it needed to maintain hightransaction volumes, FoxMeyer signed a contract withUHC, whose business at that time was not a good fitwith FoxMeyers. According to our model, this is astrategic change, which has a major impact on the firm,but which occurs infrequendy (Relationship 1).

lo handle such a change effectively requires adjust-ments at lower levels of the model (Relationship 2).Were appropriate adjustments made? The ans-werappears to be no. The response to the increased com-plexity resulting from FoxMeyer's UHC contract was toshorten the implementation schedule by 90 days, withthe result that there was no time to reengineer businessprocesses. Even more significantly, the team tested onlythose parts of R/3 that had been configured, apparentlywithout realizing the integrated nature of the softwarecould, and did, lead to errors in the rest of the system,errors that proved costly to the company.

What should have happened? One of twoapproaches could have been taken to address FoxMey-ers changed circumstances, both of which representcorrections at the lower project level (Relationship 2).First, the scope of the project could have been reduced.A decision could have been made, for example, not toroll out the package to the UHC warehouses in the ini-tial implementation, or to postpone the automatedwarehouse system until the ES implementation was

complete. Second, the implementation time could havebeen extended to ensure the system was working effec-tively prior to rollout.

Why were the appropriate steps not taken? Thereappear to be two major reasons. The first, and mostobvious, reason is that project oversight, project man-agement, and the project methodology appear tohave been ineffective. It is unclear whether there wasan appropriate oversight body at a sufficiently highlevel within the firm monitoring project perfor-mance. Further, it is interesting to note that we couldfind no mention in our information sources of whoactually managed the project, which may indicatethat the responsibilities were unclean No matter whowas responsible, project leadership was apparentlylacking, and management of the project was ques-tionable.

The second reason lies in FoxMeyer's culture,which did not invite open communication and there-fore information sharing. Even when IS employeesbelieved that incorrect decisions were being made,they were not forthcoming. Managements belief, asevidenced in statements from the projectchampion/sponsor, Anderson, that technology couldsolve any problem, may have deterred employees frompresenting a realistic view to management, which, inturn, may have precluded FoxMeyer from addressingimplementation issues more effectively, "We weregiven an assignment to find any gaps in the SAP sys-tem," recalls one FoxMeyer information systems man-ager. But systems people found they were encouragedto minimize problems. It wasn't appropriate to criti-cize SAP," the manager says. Adds a consultant whoworked on the project: "Eveiy time we showed some-thing that didn't work, they'd say, 'Is this a deal-breaket?' Well, no one was a deal-breaker. But if youput enough cinder blocks in a rowboat, it's going tosink" 12].

A question that inevitably arises from this case iswhether the FoxMeyer debacle was a failtire of technol-ogy or of management 110]. R/3 had not, at the time,been used for large volumes of transactions, and whenimplemented succeeded in handling only 10,000 trans-actions per day, compared with the 420,000 transac-tions that FoxMeyer was processing on its mainframes.Hence the implementation could be seen as a failure oftechnology. However, this technology faux pas was theresult of a management error: simulations were runwith insufficient data [2]. Further, decisions regardingtechnology are made at quite low levels in the "organi-zational change" structure of the firm as depicted inFigure 1, demonstrating that management had ampleopportunity to make appropriate decisions regarding itsuse. Hence, the error with regard to the choice of tech-

COMMUNICATIONSOFTHEACM April 2002/Vol 45 No, 4 79

nology is one of many management errors. As an exter-nal validity check of the basic structure and implica-tions of our model, we offer the fact that the majorityof project failures do not stem from technology issuesper se but from management issues surrounding theimplementation (see, for example, [4]).

Model-based Analysis of Dow Comings Abilityto Withstand Environmental Change. Havingaddressed the biggest "change" in these two cases, thatof a change in FoxMeyer's external business environ-ment, which occurred when its R/3 project was alreadyunder way, and having attributed the problems inaddressing it to lack of effective project management(Relationship 2), let us examine whether rhe sameproblems were likely to have resulted had the environ-mental change occurred at Dow Corning rather thanFoxMeyer.

First, in terms of our hierarchy of organiziitionalcontext effects, we again note the importance oi orga-nizational culture to these two implementations. WithDow Coming's open culture, employees would mostlikely have alerted management to the possibilities oifeilure if the company had decided to forego system

Lessons Learned

What are the lessons to be learned from our

model and its application to these two cases?

• Respond to environmental and strategic changes

at more tactical (project) levels.

• Ensure you have a well-defined business vision and

a strategy by which to achieve it; this may com-

pensate for not doing other things quite so well.

• Recognize the importance of organizational cul-

ture; foster an open culture and encourage open

communications.

• Take a realistic view of the role technology can

play in supporting your firm's strategy; engoge in a

strategy of "small wins" to leverage knowledge

gained.

• Manage the project; employ a strong project leader

and a well-defined methodology, so that changes

during the project are addressed appropriately.

• Do not underestimate the complexity of converting

data and creating interfaces to other systems.

• Be flexible in adjusting to unforeseen complexi-

ties; compensate at the project level by deferring

the Go Live date, reducing the project scope,

changing the number/composition of teams, orga-

nizing training, think tanks, and so forth.

• Manage implementation risks; develop contin-

gency plans and measures to assess when ta

invoke. U

testing. Contrast this situation with that at FoxMeyer,in which employees were reluctant to approach man-agement.

Second, we have seen that both Dow Corning andFoxMeyer engaged in very little project planning, didnot have good project methodologies, and were drivenby deadlines. However, there is a major distinction inthe way that Dow Corning and FoxMeyer treateddeadlines. When it appeared that Dow Corning wouldnot meet its deadlines for its U.K. pilots, it used a dif-ferent strategy from that used at FoxMeyer: the com-pany cut the scope, rather than compromising testing.

Third, Dow Corning had an excellent businessvision, supported by well-defined, long-term projectedoutcomes. Despite a questionable approach to the proj-ect, the implementation remained on track because thecompany remained in control. 1 his is an example ofRelationship 3 in Figure 2, the fact that if certain high-level aspects of the project are handled well, the projectwill not be jeopardized if certain lower-level aspects arehandled less well.

Fourth, Dow Corning had strong leadership fromthe project champion (Lacefield), something thatappeared to be lacking in the FoxMeyer imple-mentation.

Fifth, a further characteristic of Dow Corning,which would have stood the company in good steadhad it been in FoxMeyer's situation, is that it demon-strated an ability to react to adverse circumstances. Forexample: when the project was losing momentum inFebruary 1996, it reduced the number of teams work-ing on the project from eight to tour to regain focus;when it realized that local knowledge was needed forregional implementations, the company formed areateams; when area teams could not address R/3 issueseffectively, it established the PRIDE Academy to trainlocal members; and the company eventually addressedchange management as a side effect of the think tanksset up to address future business processes.

We note, however, that there is a downside to beingreactive rather than proactive. An issue for Dow Corn-ing is whether the company ultimately would havebeen better off not to have had to learn everything foritself, but instead had learned from consultants.Nonetheless, Dow Comings global R/3 project wascompleted in December 1998, on time and withinbudget [9, part C]—see the sidebar "Lessons Learned."

ConclusionHere, we evaluated the utility of a model of risk factorsin ES implementations by applying it to SAP R/3implementations at Dow Corning Incorporated andFoxMeyer Drug Corporation. We used the model tosuggest why Dow Comings implementation suc-

80 April 2002/Vol 45. No 4 COMMUNICATIONSOFTHEACM

ceeded while FoxMeyers implementation failed. Acomplete description of our model, wiiich applies to ESimplementations in general, can be found in [11].

Inherent in our model is the fact that no one risk, onits own, may be responsible for system failure, but thatcertain risks can be overcome if certain other risks aretreated particularly well. We have taken a first step inarticulating the relationships among certain risk factorsthat lead to effective trade-offs. Investigation of furtherimplementations will enrich this knowledge. Further,we expect that, as we improve our knowledge of factorsinfluencing the successful implementation of such ESpackages, we will be able to identify critical risks notonly from the adverse viewpoint of failure, but also oftime and cost overruns. B

REFERENCES1. Adlcr, P.,S. and Shenhar, A. Adapting your technoiogica] base: The orga-

nizational challenge. Sloan Management Rei/ieu.' {Fall 1990), 25-37.2. Buikeley, W.M. When things go wrong. The WdlStreetJouniaL (Nov. 18,

1996), R25.3. Davenpon, T.H. Purcing the enterprise into enterprise systems. Hanwd

Biisimss Rnnew {]\^y-h\xi. 1998), 121-131.4. Ewusi-Mensah, K. Crirical issues in ahandoned information systems devel-

opment projects. Commun. ACM 40. 9 (Sept. 1997). 74-80.5-Jesitus. J. Broken promises? FoxMeyer's project was a disaster. Was the

company too aggressive or was it misled? Industry Week (Nov. 3, 1997),31-37.

6. Kanter. R.M., Stein, B.A., and Jick, T.D. The Challenge of OrganizationalChange: How Cnmpanies Experience It and Leaders Guide It. The Free Press,Maxwell Macmillan, 1992.

7. Kotter, J.P. Leading change: Why transform at ion efforts fail. Harvard Bttsi-ness Rei'iew 73, 2 (iMar.-Apr. 1995), 59-67.

8. Markus, M.I.. and Benjamin, R.I. The magic btiilet theoi:>' in IT-enablcdtransformation. Sloan Management Rei'ieii'38. 2 (Winter 1997), 35-68.

9. Ross, J. Dow Corning corfwtation (A): Busines,s processes and information[echnology: Dow Corning corporation (B): Reengineering global processes;Dow Corning corporation (C): Transforming the organization. Center forInformation Systems Research, MIT Sloan School of Man^ement, 1996and 1999.

10. Scott, J.E. The FoxMeyer Drugs' bankruptcy: Was it a failure of ERP? InProceedings ofthe Association fhr Injhrrnation Systems Fifth Americas Confir-mee on Information Systems (Milwaukee. WI, August 1999).

11. Ston. J.E. aiid Vessey, I. Toward a Midti-Level Theory of Risks in EnterpriseSystems Implementatiom. Working Paper, 2002.

12. Williamson, M. From SAP lo 'nutsl' Computenuortd 31, 45 (Nov. iO.1997). 68-69.

J U D Y E . S C O T T (iudy.scott@cudcnvcr.edu) is an assistant professor in

the Graduate School of Business and Adniinistration at the

Universit)' of Colorado at Denver.

IRIS VESSEY (ivessey^indiana.edti) is Professor of Information

Systems at Indiana University's Kelley School of Business.

Permission to mftke digital or hard copies of all or parr cit" ihis work fur personal or ciiiss-room use is t'ranrcd witliouE f'te provided thai copies are not made or distributed for profiior commercial advanr.i^c and chet copies bear this nance and the (uil litacion on the firsipage. To copy otherwise, tii repiiblish. to post on servers iir to redistribute to lists, requiresprior specific permission and/or a fee.

© 201)2 ACM 0002-0782/02/0400 $5.00

t'-Pg** 11—Ml— q HI

i»iJlA*d n*t-*K

Fo* FnW tfi^mj

Bad

Of

i m

k i

*•».<• <li>i-Jl, !-*——« 1

i W tipaai of Dw cvivul

OB ACH, toe Thi H V * * • «»

. . . , ^

Computing(ISSN! n i l "•?'•)

is alNE^X' bimonthly electronic publicationproviding access to high quality, archival resources suitable

for use in support of computing education. Resources

include scholarly articles with wide applicahility as well

as niultiniedia and visualization resources, lahoratory

Vmaterials and other materials of practical use

in support of learning in the computer sciences.

1531-4278239$35 Professional Member$30 Student Member$130 Non-Member(for residents outside North America only)

Contact ACM Member Services1.800.342.6626 (U.S. and Canada)+1.212.626.0500 (Global)+1.212.944.1318{Hours: 8:30am—4:30pm, Eastern Time)

acmhelp@acm.orgACM Member ServicesPO Box 11414New York, NY 10286-1414 USA

Association for Computing MachineryThe First Society in Computingwww,acm.org

AD22

COMMUNICATIONSOFTHEACM April 3002 /VQI . 45. No 4 8 1