offshoring because their systems were more amenable to the offshore model. The
Distribution IT Group deals mostly with rolling out systems for the distributors,
while the Finance IT Group does mostly packages. Both are less amenable to offshore.ā
In 2002 Mike Hudson, the new CIO, became irritated with this situation, saying
in a meeting:
āWhy are we offshoring when most of their people (the providersā™) are here
Driven by this and other factors, Hudson decided on a new change strategy: moving to
centralization, standardization, and shared services. The multiple offshore provider
approach that had been in existence by default was eliminated in favor of one primary
Indian provider, one of the Tier-1 Indian providers. IT application support became a
140 Managerial competency
centralized function. āIf you have to ļ¬ght the (offshore) battle project by project, it is
much harder. With centralized management control it becomes easier,ā argued one
When asked about soft persuasion, a manager replied in frustration, āweā™ve been
using soft persuasion. Weā™ve been using it for 10 years.ā Hudson tried a āstick
approachā by setting some offshore mandates. But, when the mandates were not
met, he did not hold the Group IT managers accountable for failing these mandates.
āPeople made excuses,ā said one source.
The change was coming from local champions. In 2003, one of General Marvelā™s
three major IT Groups, the Finance IT Group, did the least offshoring. This changed
when Ted Chung became the new Finance IT Group head. Chung was an offshore
champion. And, within 12 months, the Finance IT Group became the most active
By 2004 General Marvelā™s IT workforce stood at 3000 including outsiders. Of
this workforce, 15% were offshore in India at lower costs and another 15% were
onsite in the USA, but were employees of an Indian-based provider.
After almost 15 years of offshoring, one manager quipped: āweā™re just getting
started to change peopleā™s minds.ā
General Marvel used many of the ingredients of the change management recipe:
it implemented new organizational structures, it had successful demonstration
projects, it instituted rewards, and it fostered offshore leaders. However, none of
these ingredients made dramatic changes in General Marvelā™s organizational
practices or culture. None seemed to create an offshore momentum.
All the same, the gradual, small changes that occurred over many years may
well be a mark of success. While there was no offshore zeal, there was
considerable acceptance. Furthermore, there were no episodes of broad
employee backlash to offshoring. In late 2004 General Marvel announced that it
would open its own center in India in 2005.
Most managers who interview in the business press tend to favor the latter approach for
Governance in offshore outsourcing
Short deļ¬nitions for the notion of governance are not terribly useful, so a detailed one
is given here that will be expanded upon in the rest of this section. Governance deals
with aligning the strategies and goals of the client organization with the provider,
141 Managing the offshore transition
cascading these goals down the respective organizational hierarchies, creating appro-
priate organization structures for both client and provider to achieve these goals, creat-
ing relationships and open communication channels, creating a control and monitoring
framework, and measuring the provider performance. In sum, governance is a joint
responsibility of the client and the provider in which both parties set up and agree to
roles, responsibilities, relationships, measures, problem solving processes (escalation),
scope of work, and termination processes.7
Three principles permeate the notion of governance. First, there need to be many
communications channels between the client and the provider that are open at all times
for effective dialogue. Second is relationship management, in which there is mutual
recognition of interests, as well as trust between individuals. Third, there needs to be
constant reporting by the provider at all levels (operational, tactical, business), and at
various appropriate intervals: whether this be by the minute, the day, or the month.
Governance is an expensive part of the overhead needed in offshore outsourcing,
estimated at 5% of the outsourced contract value for domestic outsourcing and 6ā“7%
for offshore outsourcing. Due to high costs, governance structures are only instituted
above contract amounts of 50,000ā“100,000 euros and typically in engagements between
large global ļ¬rms and large providers.
The higher governance costs stem from several factors. Global ļ¬rms typically oper-
ate in multiple geographical regions and therefore this leads to greater coordination
effort in software development projects. The geographical spread also has legal impli-
cations since multiple national laws apply to infrastructure management engagements.
There are higher personnel costs since the client has to designate governance roles at
various locations and with specialized expertise. For example, international business
experience is desirable: a US-based client working with a Brazil-based provider should
try to ļ¬nd a US-based contract manager with some Brazilian experience to manage the
offshore provider. Lastly, the global ļ¬rm usually suffers from multiple standards and
multiple hardware and software platforms across the organization.
The Service Level Agreement (SLA)
The SLA is a central element of governance. The SLA deļ¬nes the contracted quality
and quantity of the services along a number of dimensions. It is the operational and
legal mechanism by which the client claims the contracted services from the provider.
Before we delve further into the SLA, we reiterate the two types of IT services that
can be offshored: software development and infrastructure management. Brieļ¬‚y, soft-
ware development refers to early life-cycle activities including design and coding.
Infrastructure management includes operational services, helpdesk, and performance
trouble-shooting. The key difference between the two revolves around duration and
interaction. Software development is a relatively short-term engagement with ever-
changing requirements that involves frequent interaction with end users. Therefore the
142 Managerial competency
key difļ¬culties are in collaboration issues such as communication across cultures. On
the other hand, infrastructure management is a long-term engagement, with relatively
limited interaction with end users; it is process driven, with fairly stable requirements.
The focus in an SLA for software development is on completion of deliverables on
time and according to the contracted requirements. For example: āThe application has
to be implemented by December 31, 2005 according to the speciļ¬cation in Appendix A.
The implementation has to be executed according to the implementation plan in
Appendix B including testing and approval of the client.ā In contrast, an SLA for sys-
tem management focuses on service hours, availability, and downtime (see Table 7.1).
Of the two types of IT services, software development and infrastructure manage-
ment, discontinuity risks are relatively low for the ļ¬rst. For example, downtime of 1ā“2
days is usually acceptable. For infrastructure management, on the other hand, risks are
high, since they involve managing critical client systems at the heart of the clientā™s
business. This means that there are high threshold requirements for the availability of
communication facilities. In cases of instability or crisis offshore, providers should be
able to transfer service provisioning to another data center in another country in order
to meet the agreed service levels.
As part of the SLA, larger offshore outsourcing relationships often use a Balanced
Scorecard for reporting and discussion purposes. This is a kind of dashboard for the client
to monitor the providerā™s performance. The indicators on a Balanced Scorecard are more
business-oriented than technical or operational. Scorecards typically have four per-
spectives: business processes (e.g., lower personnel costs by 40%, decrease inventory
costs by 15%), customer perspective (e.g., ability to add new product offering within 4
weeks), organizational learning (e.g., all employees are able to use the implemented IT
system prior to the end of the year), and ļ¬nancial perspective (e.g., increase sales by 10%).
Unlike the service levels that appear in Table 7.1, the scorecard aligns the business goals
of the customer more closely with the providerā™s. Scorecards may be combined with
performance incentives for the provider, though this has not yet become a common
practice in offshoring.
A common SLA component is the speciļ¬cation of methods to improve delivery and
reduce risks. The customer and provider choose from the āalphabet soupā of methods
for this. For software development they include: CMM, ISO 9001, and PMBOK. For
support they include: ITIL, Six Sigma, and BS 7799 security certiļ¬cation.
Governance structures and key roles
In this section we examine a generic offshore governance structure and then compare
it with an example from a global corporation to see how this structure is adapted, and why.
A proper generic governance structure appears in the diagram of Figure 7.2 along with
a relationship and responsibilities matrix in Table 7.2. There are several governance
principles that drive these displays. First, the roles and responsibilities in these large
143 Managing the offshore transition
Table 7.1 Example of SLA for support of enterprise systems services, such as SAP
Service level elements Service levels Penalty reporting period
ā— System type 1 7*24 hours Not applicable None
ā— System type 2 Europe: Mā“F Not applicable
ā— System type 1 99.8%, based on a rolling 25% of monthly Monthly
3-month average cost
System type 2 99.0%, based on a rolling 25% of monthly Monthly
ā— ā— ā—
6-month average cost
Maximum downtime, 16 hours (for System type 1 50,000 USD Monthly
cumulative per year servers), (limited to for each
inside service window production servers) additional hour
Maximum number of Bi-monthly
4 hours during the
last 12 months
ā— System type 1 2 Not applicable
ā— System type 2 4 Not applicable
Maximum number of Bi-monthly
during the last
ā— System type 1 6 Not applicable
ā— System type 2 12 Not applicable
Response time (as 90% within 1second, 25% of monthly Bi-monthly
reported by the (limited to cost
standard internal production systems)
SAP tool) 95% within 4 seconds 25% of monthly
(limited to production cost
organizations are set up in a symmetrical structure in terms of geographical location and
hierarchy. Second, multiple channels of communication are established and aligned.
Third, cascading levels of strategy, partnership, relationship, coordination, and minute-
to-minute service delivery are carefully deļ¬ned between the two sides. Fourth, is the
proximity principle. Governance structures in offshore outsourcing beneļ¬t from
144 Managerial competency
rap HQ HQ
Strategy Overall Strategy
corporate management + Global
location Overall Office
IT strategy Overall coordination
Local Local customer
Local Local customer
America operations interface
Local Local customer
Figure 7.2 Generic governance structure between client and provider.
proximity. Technically there is no need to have a local customer interface but commu-
nication is improved due to proximity.
The two most important governance units are the Information Ofļ¬ce (in case of soft-
ware development sometimes called Project Management Ofļ¬ce (PMO)) on the client
side and the Global Ofļ¬ce on the provider side. The Information Ofļ¬ce and the Global
Ofļ¬ce are the highest-level units on both sides: all disagreements escalate eventually
to these two ofļ¬ces, and on rare occasions they escalate further to the executive levels
of each side. The clientā™s Information Ofļ¬ce manages both local and offshore out-
sourcing and oversees coordination of operations in different countries.
Some large ļ¬rms have not established a robust Information Ofļ¬ce and it is therefore the
providerā™s responsibility to help them structure and build such a unit. It is in the providerā™s
interest to deal with a strong Information Ofļ¬ce because this unit tends to structure and
formalize the business requirements. The provider is also acting in its self-interest
because such a unit will enhance the relationship.
In offshore outsourcing, some clients establish a Single Point of Contact. This does not
necessarily contribute to effective governance. Effective offshore governance requires
multiple channels of communication at different levels and different locations. These
145 Managing the offshore transition
Table 7.2 Roles and responsibilities within governance structures
Information Local HQ Global customer Offshore
Responsibilities HQ client Ofļ¬ce operations provider Ofļ¬ce interface operations
IT strategy CIO
Partnership Board CEO Global
management member Sales
Relationship CIO Business Global Sales
management Information Unit Customer Manager
Managers Manager Unit
Coordination Information Business Global Local Service
Managers Unit Customer Customer Delivery
Service Manager Unit Unit Manager
Delivery Manager Manager
Service Business Process The IT Service
delivery Unit experts profes- Delivery
Manager sionals Manager
End users IT
channels are needed at the strategic relationship level (the Information Ofļ¬ce), at the
tactical level (IT and business management of the local operations), and at the service
delivery interface on the operational level (IT management of the local operations).
The client manages the provider via two key documents introduced earlier: the SLA
(at the operational and tactical level) and the Balanced Scorecard (at the strategic
level). The service reports provide evidence of the providerā™s performance relative to
the SLA. Some clients manage their multiple provider contracts as a portfolio, in order
to avoid a provider lock-in, and threaten to replace non-performing providers based on
sub-par reporting vis-Ć -vis the SLA.
On the provider side, the Global Ofļ¬ce is responsible for overall coordination of all
aspects of the clientā™s services, wherever in the world they may be. Importantly, the
Global Ofļ¬ce is situated near (or if possible at) the client home ofļ¬ce.8
Letā™s take a look an actual offshore governance structure and see how it differs from
the generic structure and why. Figure 7.3 shows the governance structure between the
large France-headquartered provider Atos Origin and one of its large global clients that
we will call CPG Inc.
146 Managerial competency
CPG Inc. is an actual US-based consumer packaged goods company with extensive
European operations and a European Information Ofļ¬ce in Belgium. By 2002, CPG Inc.
had been buying domestic outsourcing services from Atos Origin for over 10 years.
Initially Atos Origin was offering these business-critical IT services out of its Belgium
delivery center directly to CPG Inc. operating companies all over Europe and Asia.
In 2002 CPG Inc. decided to offshore some of these operations in order to reduce
costs. The services chosen to offshore in this case were infrastructure management and
included activities such as operations desk, routine changes for DNS management,
routine changes to printers, patching midrange systems, ļ¬le system management, server
reboots, and kernel parameter changes. The annual amount of the contract was about
100,000 euros, representing a small percent of the contract value of the Atos Origin
CPG Inc. relationship. The provider chose to deliver the services in a combination for-
mat from both Poland and India despite the additional communication and coordina-
tion costs. This need was acute at the contract signing-time because of the heightened
tensions between India and Pakistan. If staff in both locations were actively involved in
daily operations they could easily take over for one another in case of crisis.
The three arrows in Figure 7.3 point to three areas of governance structure where the
CPG Inc.ā“Atos Origin case differs from the generic governance structures pictured in
The ļ¬rst arrow points to a special structure in the providerā™s Global Ofļ¬ce. Since
CPG Inc. is a US-based ļ¬rm it was more effective to have the providerā™s global client
executive based in the USA rather than at the European Global Ofļ¬ce which coordi-
nated the European service provision. But this resulted in a split responsibility for the
overall relationship: the Belgium-based Customer Unit Manager was limited to the
offshore relationship, while the global client executive was responsible for the entire
relationship. This split responsibility required frequent (twice a week) conference calls
between the two.
The second arrow points to the key roles created by Atos Origin. The provider had
to implement an alignment between the Polish and Indian service delivery groups so
that in case of emergency the service provisioning could be transferred to the other
location. Another alignment was the shift work involved in 24*7 operations. The two
locations split the shift work since the cost of the night shift in India was signiļ¬cantly
higher than the day shift. Atos Origin assigned 11 full-time equivalents (FTEs) to the
contract: 7 in Poland, 4 in India, and a half-time Poland-based Service Delivery
Manager. The Polish Service Delivery Manager acted as the interface to all the clientā™s
local operations in both Europe and Asia. Most of his attention was devoted to coordi-
nating the IT professionals in Poland and India, since there were large cultural differ-
ences between the two.
The third arrow points to the two additional governance roles that the client, CPG
Inc., needed to create at its main European center: a Program Manager and a Contract
Manager. These two roles represent the main increase in governance costs relative
147 Managing the offshore transition
CPG Inc. Atos Origin
og HQ, USA Regional HQ, USA