Beruflich Dokumente
Kultur Dokumente
Experience
Author(s): Charles E. Clark, Nancy C. Cavanaugh, Carol V. Brown and V. Sambamurthy
Source: MIS Quarterly, Vol. 21, No. 4 (Dec., 1997), pp. 425-455
Published by: Management Information Systems Research Center, University of Minnesota
Stable URL: http://www.jstor.org/stable/249722 .
Accessed: 26/01/2015 12:34
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Management Information Systems Research Center, University of Minnesota is collaborating with JSTOR to
digitize, preserve and extend access to MIS Quarterly.
http://www.jstor.org
MISQuarterly/December
1997 425
426 MISQuarterly/December1997
Wiersema 1995). By the end of 1993, in antici- systems development responsibilitiesfor their
pation of this paradigmshift, CEO Smith had business clients. Four transformationalgoals
launched several significantinitiatives,includ- were identified:
ing a majorculturalchange and a new role for
ITand the IS organization. 1. Move from a legacy, mainframeapplication
developmentenvironment,characterizedby
long development cycle times and exces-
sive maintenanceactivities,to a more busi-
IS transformation
thrust ness responsive environment, character-
ized by short development cycle times, a
Historically,the IS organizationat Bell Atlantic philosophy of continuous product innova-
was designed to providesupportfor products, tion, and delighteduser clients.
services, and operationsof the companywithin
its regulatedmarkets.An internalsurvey of Bell
Atlantic'ssenior management revealed that the 2. Infusestrategicthinkingwithinthe IS organi-
IS organizationwas viewed as playinga "neu- zation to conceptualize and execute novel
tral"role. When RalphSzygenda took over the business applicationsthat sustain the firm's
reins of the IS organizationas CIOin 1993,2he competitiveagilityand customersatisfaction.
found an IS workforce whose expertise was
concentrated in large mainframeapplications. 3. Buildan ITtalent base that providesa solid
The prevailingcapabilitieswere appropriatefor internal pool of skilled IS professionals,
stable, regulated,and protectedbusiness envi- trainedin currentskills (client/server,object-
ronments: long (two-plus years) development oriented,networking),customer relationship
cycle times, a narrowtechnologicaltalentbase, management, and project management.
and a low strategicfocus in ITapplications. Further, create a process for the rapid
absorption of new skills and retirementof
The hiringof a new CIO with responsibilities obsolete skills.
for IT and business reengineering signaled a
fundamentalchange in the role of the IS orga-
nization:from a reactive, cost-controlenabler 4. Buildan IS workforcewhose values change
to a proactive, strategic differentiatorof the from an entitlementmentality("thecompa-
business. Faced with the corporate transfor- ny is responsible for my career progres-
mational business pressures and the new, sion")to an entrepreneurialmentality("Iam
emergent role of IT, the buildingof a change- responsible for enhancing my skill and
career opportunities").
ready IS organization became an important
priorityon Szygenda's agenda:
To help achieve the IS transformationalobjec-
As a companywe need to be agile,dynamic;
we haveto be ableto movequickenough.... tives, Szygenda also deliberatelyencouraged
his management to "experiment" with innova-
First,I hadto havethe best team possible.I
neededpeoplewho hadbeen therebefore- tive organizationaldesigns.
winnerselsewhere. Second, we needed to
create the environmentfor a rapid-moving The most significantof these initiativeswas a
radical redesign of the Large Business and
company,withacquisitions, and
divestitures,
alliancesovernight. InformationSystems (LB& IS) unitwithincorpo-
rate IS, implemented in July 1994, under its
At the end of 1993, the IS organization was visionary,Charles E. Clark,vice presidentand
restructured(see Figure 1): five direct reports IS-LB&ISCIO.This unit of about 250 employ-
served as group informationofficers for the ees is responsiblefor new applicationsdevelop-
restructuredlines of business (LOBs)and had ment and maintenancefor three LOBs located
withinthree majorgeographicclusters. Priorto
Clark'sarrivalat Bell Atlanticin January 1994,
2Szygendaresignedthis positionto become CIOat General
Motorsin June 1996. IS-LB&IShad done a fairlygood job of staying
MISQuarterly/December
1997 427
CO
(0
Cs
CD
-i
---L
Ralph Szygenda
CIO
Vice President, IS& CIO Vice President, IS & CIO Vice President, IS
Billing Systerms 1 Enterprises, International Business Process Reengineerirng
r-
Vice President, IS & CIO Vice President, IS & CIO Vice President
I Cornsumer & Small Business
F large Busintss & Information Services Information Prcxessing
-I
Vice President, IS & (C0I Vice1Pnsident, IS Vi<KPresidentl
Corporate Systems NNtetwork Information T(ec nology
I
ahead of the demand curve and reducingunit design in mid-1994,the IS-LB&ISunit has wit-
costs. However, its resources were primarily nessed significantimprovementson a variety
devoted to the supportof legacy systems that of measures: customer satisfaction, projects
wouldsoon outlivetheirusefulness. In addition, completed on-time, and the resident IT talent
the IS managers were responsiblefor a variety base (see Table 1). Commentsby senior client
of activities,includingapplicationsdevelopment managers on formal IS project team evalua-
and maintenance, defining new projects and tions provideadditionalqualitativesupport:
enhancements, managing customer relation-
The productsprovidedare high in quality,
ships and personnel resources, and developing and responsiveness.Andall at a
the careers of IS personnel. Generally,these innovation,
tag"withinbestcost.
"price
managers' first prioritywould be applications
developmentand maintenance,leavingthem lit- [The IS team is] alwaysthinkingaboutthe
tle time to focus on strategicthinkingand cus- impacton customers: andexternal.
internal
tomer relationshipsmanagement, let alone the
development of new skills for IS human IS met everydeadlineto date on the project
resources. The existing type of managerial andhas exceededexpectations on the quality
focus worked well in a structuredand stable andthoroughnessof thedeliverables.
legacy environment, where the applications I have founda significantimprovement over
were at least five to 10 years old, and the work 1994-I amverypleasedwiththe team'ssup-
force was rewardedfor subject matterexpertise
portof [the application]this year-job well
and consistent, repetitive performance. done!!
However,such an organizationwas not appro-
priatefor the emerging role of IS as a strategic One client manager provided this assess-
differentiator of the business. According to ment of the CoE design impact on IT talent
Clark,"toparaphrasea Chinese proverb,it was leadership:
realized that if we didn'tchange our direction,
we mightend up 'wherewe were headed.'" OurISteamis moretechnically than
qualified
theirequivalentsat [anoutsidesoftwarecon-
tractor].Theirtechnicalexpertisehas called
the [outsidesoftwarecontractor]to account-
ability.Because our IS team is competent,the
The Change-Ready IS risksof aggressive timelinesare minimized.
Organization Design at The CoE initiative was also publicly credited
IS-LB&IS with the rapid delivery of an innovative applica-
tion. The members of this project team
The vision for a change-ready organizationat received the highest corporate honor: a Bell
IS-LB&ISwas a skills-based "centersof excel- Atlantic Spirit of Excellence Citation award. In
lence" (CoE) approach, with a deliberate the words of the project manager:
design molded after a quest of more than two
years by Clarkas an IS manager in a different In the traditionalmanagement structure an
firm.The CoE organizationdesign at IS-LB&IS employee must impress, throughjob perfor-
created distinct roles and processes for con- mance or otherwise,his immediatesupervisor
in orderto progressto jobs of more responsi-
ceptualizing strategic IT applications, rapidly
bilityand visibility.The problemwiththis struc-
deliveringITapplications,and buildingthe req- ture is that more often than not a person's
uisite ITskillbase.
supervisor becomes a barrierto and not a
facilitatorof career development.Rarelywilla
In April1994, Clarkenlisted one of his current
supervisor,after the initialhiring,examine a
directors,Nancy Cavanaugh, who had experi- person'sresumeand ask the basic questionof
ence and interest in organizationand person- careerdevelopment:Based uponthis individu-
nel development to serve in a key role as a al's skills,education,and performance,what is
CoE directorresponsible for implementingthe the best position for this individual? . . . From
model. Since the implementationof the CoE the manager'spointof view, they are respon-
1997 429
MISQuarterly/December
sible for deliveryto the clientand the develop- offeringthe highest value products and ser-
ment of people that workfor them is not a pri- vices of any competitorin the business [from
orityand, manytimes, ignoredcompletely. the April1993 Investor'sReference Guide].
This was my particular problem as well. I
Therefore, IS-LB&IS initiated a change-readi-
joined Bell Atlanticin 1985 and even though I
was a good performerand got good raises ness functional strategy with three thrusts:
and bonuses, I languishedin a series of dead-
end positions.Myresume was never lookedat 1. Rapid development of "bite-sized" applica-
and the question was never asked. Then in tions to dramatically reduce cycle times for
1994 as Nancy Cavanaughand others in the applications delivery and facilitate speeding
Clarkorganizationwere establishingthe CoE, of products to markets. According to
I started to get some attention.The question
was asked and the obvious answer was that I Szygenda, "We have to build systems in six
had been under-utilized.... months; that's driven by the business."
430 MISQuarterly/December1997
Structure:
groupsand integrator from other CoEs based on the project's skill
roles requirements.The ideal team is appropriately
skilled and tightly focused on delivering the
Structure is the manner in which people are applicationon-time,withincost, and according
to client expectations. The team is disbanded
grouped together, their roles and reporting at the conclusion of a new development pro-
relationships, and their task assignments.
Structures should be designed to be flexible ject. Most members are assigned to a different
and learning-enabling,as well as effective in project,or a trainingand development stint; a
small subset is assigned to an applicationsup-
promoting efficiency and cost-effectiveness
(Bahrami 1992; Nadler et al. 1992; Tushman portteam to ensure knowledgetransfer.
and O'Reilly1996). As shown in Figure3, the CoE design also uti-
The CoE design includes two types of structur- lizes three types of integratormanager roles:
al overlays (see Figure3): team-based group- skill center managers (CoE managers), man-
ings and integrator roles (Galbraith 1995; agers responsible for business client relation-
Mohrmanet al. 1995). The three key team- ships (account managers), and managers of
based structures are examined in more detail the deliveryengagement teams (deliveryman-
here. agers). These managers reportto one of the
five directors on the IS-LB&ISmanagement
The IS-LB&ISmanagement team, consisting team. Each one of these integratorroles is dis-
of CIOClarkand his five directors,is the steer- cussed below; a summaryof their role respon-
ing committee responsible for achieving the sibilitiesis providedin Table 3.
functionalstrategy of the IS-LB&ISunitand its
ongoing conduct. It is also the planningteam CoE managers (CMs) manage one or more
for the IS transformationand the decision body skill centers and have three primaryresponsi-
for interface issues involving other corporate bilities: (1) assign CoE personnel to specific
IS units. application projects, (2) coach, groom, and
counsel CoE members to achieve high skill
The skill centers (CoEs) are semi-permanent levels throughtrainingand development, per-
teams of technical specialists, or people formance appraisal, and career management
trainedin a specific ITskill. Each skillcenter is advice, and (3) develop rules, tools, and
a "virtual homeroom" of IS personnel who processes for the CoE. The CM position is a
share a high level of proficiencyin the specific clear signal of commitmentto a separation of
skillset.The initial12 skillcenters were defined people development work from project man-
using three overarching guidelines: (1) avoid agement and delivery: CMs are exclusively
skillsets that are too fine-grained,(2) keep the responsible for development of the CoE
total number of members at or below 30, and humanresources.
(3) create skillsets needed for the future,even
if there are no IS personnel to populate them CMs also have the sole resourcingauthorityfor
currently.Groupingemployees by skills was a all IS-LB&IS personnelbelow the level of direc-
major philosophical change: based on their tor;based on skillsets requestedby the delivery
current skillsets, all IS personnel below the manager,they assign specific CoE membersto
rankof directorselected a skillcenter to be ini- the projects.In makingsourcingdecisions, the
tially assigned to and a CoE that they would CM weighs an individual'sneeds for training
like to belong to in the future. (For specific and developmentagainst a project'sneeds for
skillsets, see people skills below.) expertise. Since the projectteams are tempo-
raryassignments, the CM is also the "perma-
Delivery engagement teams are temporary nent report"for an employee and provides a
projectteams created for each new approved continuityto the counselingrole.
development(or support)project.These teams
are formedby drawinga deliverymanagerfrom CMs also share responsibilityfor maintaininga
the DeliveryManagementCoE and individuals state of IS workforcereadiness;they anticipate
432 MISQuarterly/December1997
Structure
i~~ I
i. ~ People Skills
^_^_^,^^^_ ~Rewar
I~3
N
-Fiure2. StarModel Components
CD Fiure
v. 2. Star Model Components
I /^ ( IS-LB&IS >
/\ Management Team
3
/
CD /
elivery Engageme
\( Centers of Excellence Teams
X
0]Role
CM = CoE Manager
A,/,- A
.....X4/- -tiAAUUIIL .
r-lVl- ivil . Inger
DM = Delivery Manager
changes in demand for current skillsets and resources assigned to the project, and (3)
plan for acquisitionof new skillsets. Trainingis deliver a quality project on-time (within six
the linchpinfor buildinga change-ready corps; months) and withinbudget. The DM's primary
beginning in 1995, the IS-LB&ISgoal for each focus is upon the delivery of "bite-size"appli-
technical specialist was a minimumof 10 days cations, while the AMs are accountable for
"skillenhancement time." This was a major customer relationships,and the CMs manage
change in human resource management: in the IS humanresources.
the past, the focus was on "justenough train-
ing"to do theircurrentjobs.
Account managers (AMs) manage the Processes
accounts of one or more business unitby part-
neringwiththeirbusiness clients in anticipating Contemporaryfirms utilizethree key process-
strategic opportunitiesfor IT applicationsand es for sustainingtheir strategic focus and per-
by serving as the client's primarypoint-of-con- formance (Ghoshal and Bartlett1995): (1) an
tact for the corporate IS organization. entrepreneurialprocess for sustaining creativi-
Accordingto the CFO of one LOB,"theAM is ty and entrepreneurshipamong front-lineman-
my window into IS. Now I have a clear face agers, (2) an integrationprocess for building
and clear interface into IS."Althoughthe AMs and integratingcompetencies across internal
have a formalreportingrelationshiponly to IS- organizationalboundaries, and (3) a renewal
LB&IS,their partnershipswith clients led them process for continualrevitalizationof ideas for
to "feel"a dotted-line reporting relationship drivingthe business forward.These processes
withbusiness management. They workwithout nurturelateral relationshipsamong organiza-
directstaff support. tional managers and collaborative problem-
solving behaviors vital to the building of an
The AM is responsible for initiatinga "state- organization's reservoir of knowledge and
ment of work"with the business client, and skills (Galbraith1993; Lawler1996; Leonard-
using this document to monitorand communi- Barton 1995). The IS-LB&ISchange-ready
cate the projectstatus to clients and IS direc- design includes three new formal processes,
tors on a regularbasis (see processes below). one of each type, as described below.
The AMis also responsible for getting signoffs
for each project phase, workingout any client Statement of work. AMs utilize an entrepre-
neurialprocess for conceptualizingvalue-creat-
relationshipproblems encountered duringthe
project,and capturingformalclientfeedback at ing applications for their LOBs. The process
the close of a project.Gettingtimelyclosure on involves collaboratingwith business clients in
the initialrequirementsphase of a projectis a order to identify innovative ways in which IT
criticalAMtask. could facilitate new products and services,
reengineeredwork processes, or reconfigured
The AMs also contribute biweekly "radar relationshipswithexternalcustomers and other
reports"to communicatethe currentand future stakeholders. As managers of customer rela-
needs of their clients to other AMs, CoE man- tionships,AMs are not only heavily involvedin
agers, and the IS-LB&ISmanagement team. the conceptualizationof innovativeIT applica-
These reportsare importantfor anticipatingthe tions, but also in defining business require-
human resource skillsets needed for future ments, scheduling prototypeevaluations, and
projects. planning roll-outs. The outcomes of this
process are documentedin a statementof work
Delivery managers (DMs) manage projects (SOW)which becomes a livingguide for work
for new applications or provide support for by the deliveryengagement teams. The SOW
existing applications. Their responsibilities is a criticalmechanismfor linkingthe activities
include: (1) identifyskills needed for delivery of the AMand DMroles, because the DMs are
engagement managementteams based on the responsible for delivering an application
statement of work, (2) manage the team of IS accordingto the SOWdeveloped by the AM.
436 MISQuarterly/December1997
1997 437
MISQuarterly/December
Lawler 1996). Reward systems that attract, teria. First, the change-readiness functional
develop, and motivate the rightindividualsfor strategy (rapid development of "bite-sized"
the new organizationalcapabilities are key to applications, strategic focus, and IT talent
the success of a new design (Gupta et al. leadership) is in alignment with the business
1992; Shuster and Zingheim1992). goals for competitiveagility, customer intima-
cy, and operational excellence. Second, the
The IS-LB&ISmanagement team felt strongly
structures, processes, people skills, and
that an appraisal process that placed high rewardsystem components were designed to
value on customer satisfaction,as well as suc- enable the change-readiness functionalstrate-
cessful projectteaming, was a criticalcompo-
gy. Third,these components are both internal-
nent of the new design. The performance
ly consistent and mutually reinforcing. For
appraisalsystem implementedonly six months example, skill centers build and nurturecom-
intothe new CoE design initiativedifferedfrom
petencies requiredfor the new integratorroles
the standard corporate IS process in several
(e.g., account management). The resource
ways. First, it included performance ratings planning process facilitates coordination
from multipleIS bosses: the skill center man- across the new team-based groups and inte-
ager (CM)and all projectmanagers (DMs)dur- grator roles. A new rewardsystem motivates
ing the prior 12-month period. Second, it leveraging of IT skillsets and teamwork to
included a process for capturing business meet changingclient needs.
client evaluationsfor each projecton whichthe
IS individual had worked. Third, the ratings Nevertheless,the CoE design describedabove
from these three sources were weighted, with was the result of a multi-yeartransformation
much higher values placed on client and DM process thatwas not totallypre-plannedand not
assessments than those of the CoE coach withoutsome unexpected pitfalls. In the next
(CM). section, an analysisof the key learningelements
of the transformationjourneyis presented.
Since the corporate human resources depart-
ment had been considering the merits of a
multi-inputappraisalprocess, the IS-LB&ISunit
was able to design and implement this new
system as a pilot for the organization.Clark's Managingthe
directorfor the CoE design (Cavanaugh)was TransformationJourney
given the responsibilityfor designingand imple-
mentingthe new system. Likeother corporate The CoE design involveda radicalreconfigura-
IS annual evaluations, the multi-input tion of structures,processes, people skills, and
appraisalsfor the IS-LB&ISunitwere intended reward systems. Successful implementation
to be used for annual raises, bonuses, and per- was dependent on the buy-inof both business
sonnel developmentplanning. and IS personnelto not only the change-readi-
ness goals, but also to the uncertaintiesasso-
ciated with an as-yet-unproven design.
Further,the transformationof the IS-LB&IS
Congruence among five organizationwas initiatedin anticipationof a
components new business environment where change-
readiness IT capabilities would be impera-
Three criteriaof "fit"have been suggested for tive-i.e., no crisis was currentlyat hand. As a
evaluating the effectiveness of organization result, an additionalchallenge was to create a
designs: (1) strategic fit, or whetherthe strate- sense of felt-need and urgency to mobilize
gy is appropriateto the competitive environ- commitmentof relevant stakeholders (Nadler
ment, (2) strategies, and (3) internalorganiza- and Tushman 1995).
tional fit, or whether the architecturalelements
are congruent (Nadler and Tushman 1992). Traditionalmodels of organizational change
The CoE design appears to meet all three cri- (e.g., Lewin 1951) suggest that the change
438 MISQuarterly/December1997
process should be plannedfor and deliberately Dean 1982). The key stakeholders for this
managed. Based on a vision of the need for design change were not only the IS-LB&IS
and natureof the change, managers should (1) managers and professional staff and IS-
sell the vision to key stakeholders, (2) imple- LB&IS'sbusiness clients, but also the rest of
ment the change, and then (3) institutionalize the corporateIS organization.
the new order(Beer et al. 1990; Goodmanand
Associates 1982; Nutt 1986; Srinivasan and The buy-in of the corporate IS organization
Davis 1987). Changingthe attitudesand values was achieved in two ways. First,organization
of organizational members affected by the design experimentation was encouraged by
change is a key partof this process. Szygenda; this provided IS-LB&ISwith the
autonomyto innovate. Second, Clark'sknowl-
However,recent researchers have argued that edge about CoE approaches and organization-
an alternativemodel that takes account of both al roles, gained while workingat anotherorga-
the anticipatedchanges and the unanticipated nization,were recognized as assets duringhis
reactions and adjustmentsis more reflectiveof own hiring process. Soon after his arrivalat
good practice (Mintzberg1987; Orlikowskiand Bell Atlantic, he presented the CoE design
Hofman1997). Anticipatedactions are deliber- concept to the corporate IS management
ate and preplanned.Unanticipatedactions are team. He argued that the IS-LB&ISunit was
innovationsthat arise as the organizationbet- large enough in personnel and range of
ter understands the complexity of a given responsibilitiesto be a "microcosm"of an IS
organizationalchange and its consequences. organization.Further,under the January 1994
restructuringof the corporate IS organization
The two-yeartransformationjourneyembarked to align with the overall firm (see Figure 1),
on by Clarkand his managementteam reflect- Clarkhad the authorityfor autonomouslyman-
ed this improvisationalapproachfromthe start. aging applicationsdeliveryfor fourLOBswithin
Clark's intent was to not over-plan the CoE IS-LB&IS.In March1994, he received corpo-
design implementation; only a few months rate IS approval for moving forwardwith his
were devoted to establishingthe new structure vision of the CoE model and implementingit
and "selling it." However, Clark had also withinIS-LB&IS.
learned through experiences at another orga-
nization that a "shallow end approach" to
entering the pool-i.e., an incremental
approach-was not an effective implementa- Use a "Jump-In"Approach
tion tactic: the CoE design was a radical
change, and the "fit"among the differentcom- With top management support behind him,
ponents of the design was one of its strengths. Clark took a "jump-in"approach: a new IS-
The challenges faced, and "lessons learned," LB&ISmanagement team was formed and a
categorized according to the three phases CoE director (Cavanaugh) was selected to
suggested by the traditionalchange models spearhead the implementation. Many of the
are discussed below. However, Table 4 also new account manager, CoE manager, and
depicts these milestones on a timelinethat dis- deliverymanager roles were filledwith internal
tinguishes between preplanned actions and candidates. In July 1994, the new structure
unanticipatedinnovations. was presented to the rest of the IS-LB&IS
staff, IS professionals were assigned to skill
centers, and the IS-LB&ISorganizationantici-
pated workingthroughthe remainingissues as
Phase 1: establish the need for they arose. The IS-LB&ISmanagement team
change and sell the vision acted as an oversightgroup,interveningwhen-
ever barriersimpeded the pace or success of
Radical and anticipatory organizational the transformation process. The "jump-in"
changes requirethe commitmentof key stake- approach helped create a sense of urgency
holders affected by the change (Goodmanand and an environmentof excitement and energy
MISQuarterly/December
1997 439
440 MISQuarterly/December1997
MISQuarterly/December
1997 441
1997 443
MISQuarterly/December
ing, while others who had alreadybeen trained expected a supervisorto create opportunities
in new technologies, such as PowerBuilder, for them. A variety of human resources pro-
were able to keep moving on to new assign- grams (newsletters,training,etc.) were devel-
ments. Since the new development projects oped in partnershipwith the human resources
were concentrated in one geographic area, departmentto help the CMs in moving more
some remote staff also felt they were being reluctant IS-LB&IS employees toward an
overlooked. entrepreneurialset of values.
In response to these problems,transitionplans Overall, the IS-LB&IS management team
for specific individualsbegan to be developed learned that not all employees embrace
as part of the transition planning for legacy change enthusiastically.Accordingto Clark:
applicationsbeginningin the fall of 1995: these Convertsmustconvertthemselves,andsome
plans were developed by the CMs responsible
for employees in the legacy skill CoEs, were may never move. Not everyoneis excited;
you won't get everyone to change with
reviewed with DMs of the legacy applications, you.... Also,peopleare at different
pointsin
and then communicatedto the affected individ- theirown life-cycles.Youneed to accommo-
uals. As IS staff began to be transitionedout of datethis,orsomegoodpeoplewillbailout.
their legacy roles without the initial fears of
their delivery managers being realized, the
concept of customized transitionplanningfor
individualstook hold, and these morale prob- Implementand Fine-Tune New Appraisal
lems subsided. Systems
Towardthe end of 1995, a less formalmecha- As described earlier,a new appraisalprocess
nism was also implemented to help control based on assessments by multiple IS man-
rumors and allay fears: brown-bag sessions agers as well as business clients was consid-
with the CoE director at the two sites geo- ered an importantcomponent of the new CoE
graphically removed from headquarters. The design. The concept of client evaluations of a
CoE director provided information "straight projectteam was new at Bell Atlantic,as were
fromthe horse's mouth,"heard individualcon- the heavy weightings on non-IS input. Initial
cerns, offered personal advice/counseling in weightings were 40% client(s), 40% DM(s),
one-on-one discussions, and relayed relevant and 20% CM. These were revised to 50%
concerns to the respective CMs forfollow-up. client(s), 40% DM(s), and 10% CM input in
mid-1995 in order to promote customer-orien-
Another continuing challenge has been the tationamong the IS staff.
"bench-time"anxieties of the technical staff:
some individualswaiting to be assigned to a However,problemsbegan to surface. In partic-
new projectfelt that they were vulnerabletar- ular,IS staff who received less-than-satisfacto-
gets for downsizing. While efforts were made ryteam ratingsfromclients expressed tremen-
to reduce bench-timethroughbetter coordina- dous anxiety relativeto their job security and
tion in the resource planning process, these promotionprospects withinnot just IS-LB&IS,
perceptions of vulnerability remain an unin- but also the IS organizationas a whole. Rather
tended impact of the CoE design in a hyper- than motivatingall team members to be more
competitiveindustrycontext. customer-oriented, the unintended result for
some IS staff was a perceptionof having little
Finally,the IS-LB&ISemployees were not uni- personal control over their own performance
formlyreceptive to the transformationfrom an ratings.Two aspects of the process were the
"entitlement"to an entrepreneurialmindset in primarycauses: (1) the business client ratings
which employees take more responsibilityfor were at the projectteam level, not the individ-
their career development. According to ual employee level, and (2) since the client rat-
Cavanaugh,"Some employees still expected a ings were collected at year-end, there were
sports agent rather than a coach"-they concerns about unfair client biases due to
444 MISQuarterly/December1997
MISQuarterly/December
1997 445
446 MISQuarterly/December1997
Lesson Rationale
Use a "jumpin"approach A jump-in-the-pool approachcreates a sense of urgency
the CoE design concept and can help sustain commitme
Don'twaiverfromthe vision The CoE visionrequiresa majormindsetchange on the p
Gainingtheirbuy-inrequireseducationand "earlywins"t
determinationto not waiverfromthe visionof the CoE de
business is crucial.
Createand nurturea team-managedcollaborative The CoE modelcreates new interdependenciesamong m
environment theirroles and responsibilities.Success becomes depend
environmentthatvalues teamingand diversityas oppose
managementstyle.
Expectnew formalcoordinationmechanismsto evolve withinthe CoE design creates signifi
Role differentiation
and alignmentof activities.New formalcoordinationmec
communicationand interpersonalrelationshipsamong in
Use contractorsto transferbest practicesfornew Contractingexternalpartnersfor projectworkdue to thei
technologies technologiesmay reducethe opportunitiesfor developin
create moraleproblems.Use contractorsto transferbest
undershort-termcontracts,not do the projectwork.
Fine-tunenew performanceappraisalsystems to A multi-inputperformanceappraisalsystem-with inputs
reducedysfunctionaleffects deliverymanagersand skillcenter managers-rewards t
design. However,dependingon the newness of a multi-
whichcomparisonsare made across IS units,accommod
to balance individualand organizationalneeds. Further,i
accountmanagementand coaching-without systems de
established,accommodationsmay be needed to ensure
not undervalued.
Manageemployee anxieties Individualsreactdifferentlyto radicalchange because of
stages. A successful transitionprocess accommodatesth
CD
Empowermentworkswell when employees are eager to
()
ca own skilldevelopmentand careers. Since not all employ
customizedtransitionplanningand special humanresou
a
o-
move the reluctantemployee towardthis goal.
(0
Communicateprogressand demonstratevalue with Stakeholdersneed proofthatthe organizationalchange i
N
performancemeasures be identified,progresstracked,and resultscommunicat
stakeholdersas earlyas possible.
4-,
ii
The analysis in this paper focused on three well as the enterprise-wide leveraging of
processes: one anticipated but not yet fully expertise and solutions.
developed (CoE life cycle management) and
two formal processes not fully anticipated The CoE design described here combines
some of the advantages of both the life-cycle
(statement of work,resource planning).As dis-
cussed earlier,the frameworkof entrepreneur- and the SBU team approaches (see Table 7).
ial, integration, and renewal processes Similar to the SBU team, the CoE design
addresses the need for IS staff to develop a
(Ghoshal and Bartlett 1995) appears to be
useful for identifying the differing thrusts of deep business understandingas well as strong
these new processes. Moreproblematicat Bell client relationships. However, in the CoE
Atlanticwas the attempt to redesign a corpo- design, this role is played by an account man-
rate-wide performance appraisal process that ager. An underlying assumption of the CoE
would nurturethe values of the CoE design design, then, is that client perceptionsof busi-
ness responsiveness do not requireintensive
(teamwork,client responsiveness, skill devel- client access to and relationshipswith tempo-
opment), yet not be perceived as misaligned
withthe corporateIS organizationas a whole. raryprojectdeliveryteam members. Since the
IS-LB&ISunit previously operated under an
How does this project-based design compare SBU team model, in which systems project
with earlier models of IS organizationdesign, team members were typically assigned to a
business unitfor long time periods,the change
specificallythe SBU-team and life-cycledesign
models? The life-cycle design sets up a struc- to a CoE design appeared to some client
turalorganizationbased on skills utilizedat dif- groups to be a step backwards.In reality,the
ferent stages of the development life cycle: design established a single-point-of-contactfor
the clientthat is totallyaccountablefor the cus-
systems analysis, technical design, program- tomer relationship and enabled the project
ming, and maintenance (Boar 1996; Zmud
managers and IS development staff to focus
1984). This approach also requires good pro- on qualityprojectdeliverables.
ject management skills and a standard
methodology to ensure integrationacross the Similarto the life-cycledesign, the CoE design
life cycle steps. The primaryadvantage here is also pools IS staff. However, the pools here
that deep IT skills can be leveraged across a were primarilyfocused not on life cycle steps,
portfolioof systems development projectswith but on new technologies to support rapidpro-
a relativelysmall pool of skilled IS personnel. totyping and new management methods. A
A majorweakness is that IS staff have limited life-cycle design approach may be more suit-
opportunities for developing deep business able for an IT skills-centricapproach under a
expertise. more stable environment. However, when IT
skills are at a high rate of obsolescence and
In contrast, the SBU team approachsets up a there is an organizationalemphasis on rapid-
structuralorganization in which IS personnel
prototypingtools, the CoE design should prove
are assigned to specific business units, func-
superiorin facilitatingITtalent leadershipwith-
tionalareas, or processes of the overallfirmin in the context of continual reskilling,in which
orderto develop a deep expertise in a particu-
today's new skillsets may become tomorrow's
lar business area and close workingpartner-
legacy skills. Similarto a consultancy model,
ships withtheirbusiness clients. Whilefacilitat- the specific CoEs implemented by IS-LB&IS
ing a greater business responsiveness, the also foster strong project management skills
drawback of the SBU team model is that it and standardmethodologieswithinand across
inhibitsa firm'sabilityto leverage IS skills and skillsets.
expertise across the enterprise.When IS skills
are expensive, in short supply, and character- Further,the CoE design model also avoids a
ized by a high rate of obsolescence, the busi- disadvantageof typicalimplementationsof the
ness-centric approachof the SBU team design two more traditionalIS organizationdesigns:
creates deficiencies in IT talent leadership as the bundlingof responsibilitiesfor projectman-
448 MISQuarterly/December1997
'1
u
tO
4b
agement and people management into a sin- resources as it waited for the deregulation
gle managerial role. This leads to inadequate landscape to unfold.
attentionto the internaldevelopment of ITtal-
ent, as short-termdelivery pressures typically Otherlimitationsto the potentialrelevance and
take precedence over long-termpeople devel- applicabilityof this CoE design are acknowl-
opment. Similarto a consultancy, the project- edged. First and foremost, the design
based CoE design sets up a separate struc- described here was only implemented within
ture for mentoringand places responsibilityfor an IS unit with applications development
resource assignments in the hands of the responsibilities; it was not rolled out to the
mentors. entire Bell Atlantic IS organization. On the
other hand, the metricsfor the success of this
This case study suggests that a CoE design transformationare all widelysought IS capabil-
based on a project-based model prevalent in ities: rapid delivery of value-added business
the consulting industryhas considerable utility applications, superior customer satisfaction,
for IS organizations in large firms outside the and the abilityto quicklyassimilate ITskillsets
consulting industry-especially in firms facing (Hofman and Rockart 1992; Merlyn and
majorindustrychanges with a long-tenuredIS Parkinson1994; Parker 1996). The key chal-
workforce relatively unskilled in newer tech- lenge to implementingthe CoE design within
nologies. However, in assessing the generaliz- only a single systems development unit was
abilityof this case study, additionalfactorsthat the institutionalizationof a new performance
may have contributedto its successful imple- appraisalprocess that would rewardbehaviors
mentation at Bell Atlantic need to be recog- associated with the change-readiness func-
nized. First,the CoE design was introducedby tional strategy, but not cause inequities for
a new corporate IS management team and membersof the IS-LB&ISunit.
was spearheaded by a newly arrivedIS execu-
tive who had prior experience with redesign Otherfactors that may limitthe applicabilityof
initiatives.Nevertheless, special effortsto gain the CoE design model includethe fact that the
client buy-inwere required,suggesting thatthe IS-LB&ISunitwas of a moderatesize (approxi-
account manager role inherentto a consultan- mately250 IS employees), and the rapidappli-
cy/client relationship may be a "hardersell" cation capabilityfor this organizationwas con-
when the consultancyand the client are partof ceived in terms of a six-month cycle time.
the same organizational"family."Second, the Project teams thus had a limited life, and IS
CoE "fit"with its organizationalcontext in sev- specialists with highly valued skills typically
eral importantways. The teamed management would be available for multiple assignments
approachfit Clark'scollaborativemanagement withina given year. For IS organizationsof a
style, as well as the Bell Atlanticvalues (e.g., different size, and with large mission-critical
teamwork) established by the CEO several projectsthat may include the reengineeringof
years earlier; the strategic focus of the new globalworkprocesses, this specific model may
account manager role fit the new corporateIT not be optimal.On the other hand, a six-month
role;the four LOBsclient units had just recent- modularrelease schedule mightyield some of
ly been formed, which set up a management the same benefits (Koch1996).
environmentconducive to mutual IS/business
adjustment;and the newly decentralizedstruc- This case study also providesvaluableinsights
ture for corporatehuman resources placed the into the process for managingthe transforma-
first-linefacilitatorsfor new CoE job descrip- tion to a CoE design. In addition to the
tions and rewardsystems under the corporate "lessons learned" (Table 5), a key insight is
IS umbrella. The timing of the CoE design that radical organizational changes must be
implementationwas also a potentiallyimpor- managed collectivelyin a teamed environment
tant factor:the transformationwas couched as and not in a hierarchical,command-and-con-
a radical redesign in anticipation of still trol mode. Radical changes cannot be totally
uncharted, major industry changes. This pre-planned;after establishing the initialmis-
meant that the organization had some slack sion and mandate, management must place
450 MISQuarterly/December1997
MISQuarterly/December
1997 451
thank the IS managers, IS professionals, and Goodman, P. S., and Associates (eds.).
business managers at Bell Atlanticwho provid- Changes in Organizations: New
ed the data for this study. We are also grateful Perspectives on Theory, Research, and
to the Advanced Practices Council of SIM for Practice, Jossey-Bass, San Francisco,
theirfundingand insights;MadelineWeiss and 1982.
Bob Zmud for their feedback at different Goodman, P. S. and Dean, J. W., "Creating
stages of this collaborative project; and the Long-Term Organizational Change," in
MIS Quarterlyeditors and reviewers for their Changes in Organizations: New
stimulating suggestions during the review Perspectives on Theory, Research, and
process. Practice, P. S. Goodman and Associates
(eds.), Jossey-Bass, San Francisco, CA,
1982, pp. 226-279.
Gupta, N., Ledford,G. E., Jenkins, G. D., and
References Doty, D. H. "SurveyBased Prescriptionsfor
Skill-BasedPay,"ACA Journal(1:1), 1992,
Bahrami, H. "The Emerging Flexible pp. 50-61.
Organization," California Management Haeckel, S. H., and Nolan, R. L. "Management
Review (34:4), 1992, pp. 33-52. by Wire,"HarvardBusiness Review (71:5),
Beer, M., Eisenstat, R.A., and Spector, B. September-October1993, pp. 122-132.
"Why Change Programs Don't Produce Hamel, G., and Prahalad,C. K. Competingfor
Change," HarvardBusiness Review (68:6), the Future: Breakthrough Strategies for
November-December1990, pp. 158-166. Seizing Control of Your Industry and
Boar, H. B. Strategic Thinkingfor Information Creatingthe Marketsof Tomorrow,Harvard
Technology: How to Build the IT Business School Press, Boston, 1994.
Organizationfor the InformationAge, John Hofman, D., and Rockart, J. F. "Systems
Wiley& Sons, New York,1996. Delivery: Evolving New Strategies," Sloan
Bowen, D. E., Ledford,G. E., and Nathan, B. ManagementReview (33:4), Summer 1992,
R. "Hiringfor the Organization, Not the pp. 21-31.
Job," Academy of Management Executive lacono, S. C., Subramani,M., and Henderson,
J. C. "Entrepreneuror Intermediary:The
(5:4), 1991, pp. 35-51. Natureof the RelationshipManager'sJob,"
Conner, K. R., and Prahalad, C. K. "A in Proceedings of the Sixteenth
Resource-Based Theory of the Firm:
International Conference on Information
Knowledge Versus Opportunism,"
Systems, J. I. DeGross, G. Ariav,C. Beath,
Organization Science (7:5), September- R. Hoyer, and C. Kemerer (eds.),
October1996, pp. 477-501.
Amsterdam,1995, pp. 289-300.
D'Aveni, R. Hypercompetition:Managing the Keen, P. G. W. Shaping the Future:Business
Dynamics of Strategic Maneuvering, The Design Through InformationTechnology,
Free Press, New York,1994. HarvardBusiness School Press, Boston,
Galbraith,J. R. Designing Organizations:An 1991.
Executive Briefing on Strategy, Structure,
Kiely, T. "InformationTechnology: Making
and Process, Jossey-Bass Publishers, San Collaboration Work, Editorial Briefing,"
Francisco, 1995. HarvardBusiness Review (75:1), January-
Ghoshal, S., and Bartlett,C. A. "Changingthe February1997, p. 10.
Role of Top Management: Beyond Koch, C. "Flippingthe Switch,"CIO,June 15,
Structureto Processes," HarvardBusiness 1996, pp. 43-66.
Review (3:1), 1995, pp. 86-96. Lawler, E. E. From the Ground Up: Six
Goldman, S. L., Nagel, R. N., and Preiss, K. Principles for Building the New Logic
Agile Competitors and Virtual Corporation,Jossey-Bass, San Francisco,
Organizations: Strategies for Enrichingthe 1996.
Customer, Von Nostrand Publishing, New Leonard-Barton,D. Wellspringsof Knowledge:
York,1995. Building and Sustaining the Sources of
1997
452 MISQuarterly/December
MISQuarterly/December
1997 453
454 MISQuarterly/December
1997
Appendix
Methodology
This case study is based on first-handexperiences of two senior Bell Atlanticmanagers leading the IS
organizationaltransformation,a multi-phasedata collectionover a periodof 14 months by two acade-
mics, and collaborativeanalyses of this case study by all fout authors. Morespecifically,this research
collaborationbegan in April1995, when the first author, the vice president and CIO of Information
Systems for the Large Business & InformationServices unit (IS-LB&IS),participatedin a telephone
interviewwith one of the academic authors for a research study on IT coordinationmechanisms in
large, divisionalizedfirms. The second author,the directorof Centers of Excellence withinIS-LB&IS,
subsequently made arrangementsfor both academic authorsto conduct on-site interviewsat three dif-
ferent Bell Atlanticsites between October 1995 and June 1996. A total of 27 Bell Atlanticemployees
were interviewed,includingthe CIO for the corporateIS organization,five direct reportsto the corpo-
rate CIO, five directors reportingto the IS-LB&ISVP/CIO,other members of the IS-LB&ISunit, plus
three business managers in the client organizationssupportedby IS-LB&IS.Numerousarchivaldocu-
ments includingannual reports,strategydocuments of the IS organization,excerpts fromkey speeches
given by the CEO and the CIO, statements of work descriptions for systems projects, job position
descriptions,performanceappraisalforms, and client feedback on completed projectswere also avail-
able to the authorsover this multi-yearcollaboration.
1997 455
MISQuarterly/December