Demystifying Outsourcing and Global Services Delivery

diginotes
Demystifying Outsourcing and Global Services Delivery

Ramesh Adavi
Sr. Consultant
Prabhas Sinha
Product Marketing Manager
Digité Inc.
82 Pioneer Way
Mountain View, CA 94041

Optimizing the SERVICES VALUE CHAIN has emerged as the most powerful lever, for improving
business performance.

This white-paper, takes a look at the important issues and drivers, of the global services industry,
and suggests a few best practices. Six Sigma, LEAN and Critical Chain project Management
concepts have proven themselves in manufacturing. They can and should be applied to services
delivery as well.

THE OUTSOURCING PARADIGM

Outsourcing is now the universally accepted solution for enhancing service delivery and
improving bottom-line performance. After accounting for transition costs and associated risks, net
cost savings of 30%-35% are achievable.

The top three business benefits of outsourcing are (source www.ventoro.com):
l 46% of the benefit value due to re-engineering business processes (a case for BUSINESS
PROCESS MANAGEMENT).
l 45% due to improvement in delivery execution (improved quality, reduced delays, shorter
time-to-market).
l Only 9% due to labor cost differences.

Surprisingly, low cost labor is not the only reason why outsourcing is here to stay

BUSINESS PROCESS MANAGEMENT (BPM)

BPM is clearly the “hottest” area in service delivery.
Points to consider are:

PROCESS VALUE-ENGINEERING

Not all processes (and activities) are “valuable”. The customer must decide the value.
v Processes must be categorized as
l VALUABLE / BILLABLE, if a customer will pay (say at an hourly rate)
l MANDATORY, those processes which are not “billable”, but considered

www.digite.com 1 Copyright 2005 Digité, Inc. All rights reserved.

diginotes

essential for administrative control or regulatory compliance
l WASTE, none of the above.
Processes such as approvals, hand-offs, inspections, are a WASTE. Unless the customer has
explicitly agreed to pay for them (wishful thinking!), or there is a MANDATORY requirement to
retain them.

Clearly, there is very little room for bureaucracy in services organizations. Overdoing the quality
bit, over-reporting, and the such-like, are symptomatic of waste, which could ironically be
rewarded!

v Co-relate SERVICE LEVELS with VALUE

l BASIC SERVICE LEVEL as per contracted terms with customer
l INCREMENTAL SERVICE LEVEL there is a perceived benefit in providing
beyond BASIC level. For e.g. incentives for delivery schedule improvement.
l DELIGHT by design, incorporate processes, which beget superlatives from
the customer. The benefits are not necessarily evident in the short term.
Nevertheless, the long term value of such processes must be qualified.

v Processes must be close to the customer.

l Value is co-created with the customer, by the “foot soldiers”. Not by
supervisors. Empowerment of customer facing teams is of paramount
importance.
l Opportunities for SELF-SERVICE processes, with an enhanced USER
EXPERIENCE, must be aggressively considered. Internet banking is case
in point. e-LEARNING is fast proving a point.

PROCESS PERFORMANCE MEASURES

For the (valuable) processes which are retained, measures must be established.
Useful process performance measures are:

v CYCLE TIME
Elapsed time, from start to finish of the process / activity / task / project.

v PROCESS CYCLE EFFICIENCY (PCE)
PCE = (VALUE ADDED Time) / (Total CYCLE TIME)
VALUE ADDED time represents, the time spent by a person on the activity for
which a customer will pay (Billable activity). Many BPOs operate at PCEs as low
as 10%. A PCE of 50% , though considered top-of-the-line, is achievable.
Substantial productivity benefits are waiting to be picked up here.

Common reasons for low PCE are:
l Waiting for work to be assigned
l Seeking clarification
l Transportation e.g. papers for signature
l Human nature
o Parkinson's Law work expands to fill the time
o Students Syndrome leave it till the last day (of the exam)

Tasks must not be allocated too much in advance, or with very big
safety margins.

v WORK-IN-PROCESS (WIP).
WIP is the number of pending tasks / activities with a person. Multi-tasking has
detrimental effect on overall throughput. Add to it the costs of frequent
changeovers between tasks. Even work quality suffers, when a person tries to
do too many things at a time.
In particular, WIP on STRATEGIC RESOURCES must be continuously tracked.
Strategic resources are those that yield high value (billable rate), and are critical
to the success of a project. Chances are that STRATEGIC RESOURCES are
heavily loaded, and are on one or more CRITICAL CHAINS.


diginotes

v SET-UP TIME.
Set-up procedures, such as PROJECT INITIATION, ACTIVITY TYPE / TASK
TYPE CHANGE OVER; have inherent penalty costs. Typically, 15-30% of
overall process costs are attributed to set-up costs. For which customers don't
pay.

v PROCESS COST

PROCESS COST = (CYCLE TIME) X (RESOURCE COST RATE*)
*Note: Only direct cost has been considered, in which, the cost of the most significant resource is
considered, typically person-time.

Ways to reduce PROCESS COSTS are:
l Reduce resource cost by outsourcing to cheaper destinations. But this is
only one of the ways

l CYCLE TIME reduction
o WASTE (and DELAY) reduction
o SET-UP time reduction
l PCE improvement

PROCESS MATURITY

Indicators of PROCESS MATURITY are:

v VARIABILTY on key PROCESS PERFORMANCE MEASURES

Immature processes have large variability on key measures such as CYCLE TIME and
PROCESS COSTS. And they in turn reflect an unstable operation!
l Deming has declared that a process is statistically stable, on a performance
measure, if its variability is within 5% of the mean. (2 ó maturity).
l Commercial / business commitments can be made, within manageable risk, only
when processes are stable. Services, compared to manufacturing processes, are
considerably more difficult to stabilize.
l FIXED BID agreements (SLAs, Projects,…), are on the rise. However, as per
Standish variability on project estimates are extremely high (69% over-run on
schedule, is the average).
l The luxury of multiplying “estimates” liberally (say by an arbitrary 2.2 times), will not
be available. Customers will not pay for “immaturity”.

v PROCESS QUALITY

The quality of a process is measured by number of DEFECTS counted in the PROCESS. A
defect is anything that does not meet pre-agreed specifications. Which points to an important, but
the often overlooked requirement, of documenting pre-agreed specifications and acceptance
criteria.

Specifications (and acceptance criteria) should be defined for different SERVICE LEVELS, with
VALUE defined for each service level. A pragmatic approach to quality is to take an end-to-end
view of quality. And not create “silos of excellence”. Also, a smart balance between cost-of-quality
and service-level is required.

The achievement of an end-to-end quality objective requires, that all stages included in the
process, perform within acceptable limits (LCL / UCL). Clear ENTRY / EXIT criteria must be set
for all key stages, and for hand-over between stages. Much like a relay race!

An estimated 25% of costs (in IT projects), are wasted as unpaid re-work, due to defects.
Damned customers! They are the guys who get to decide what a defect is! A very large
proportion of these defects are traced back to lack of clarity (and documentation) of
REQUIREMENTS.

Causal analysis techniques, are commonly applied for defect reduction:
l Pareto Principle 20% causes for 80% defects.
l Examine the correlations between defects, and factors such as process type,
location, customer, person assigned, value stream type (practice group / SBU,), etc.


diginotes

v COMPLEXITY REDUCTION
Complexity is measured in terms of number of activities, interdependencies between activities
and variety of activities (TASK / ACTIVITY TYPES), in a process.
Complex processes are also likely to be costly, because:
l Cycle-time variability increases proportionately to the number of activities
l Number of dependencies (on predecessors) increases WAIT times
l Learning-curve costs are proportional to variety.
l Set-up costs increase proportionately to variety.

ORGANIZATION MODELING

Creating the right type of “outsourcing” capacity is extremely critical. This calls for
ORGANIZATION MODELING.

The following are the prevalent outsourced organization types:

v Off-shore, staff augmentation
l Re-deploy / increase the headcount in an out-location.
v Project outsourcing
l Projects with at least 3 months duration.
l Leverage vendor's specialist skills, which are critical to the success of the project.
l In-house projects do not receive the required priority.
v Off-shore development centre (ODC)
l Vendor head-count exclusively earmarked to the corporation
l Commonly (in the IT domain), the processes outsourced to ODCs are Maintenance /
support on legacy / existing applications and Product Life Cycle Management for
product companies.
l The ODC model is expected to deliver all benefits of outsourcing process
improvement and lower cost.
l A KEY CONCERN: PROTECTION OF INTELLECTUAL PROPERTY & DATA
v Functional outsourcing
l Complexity reduction is a significant benefit of functional out-sourcing.
l Very significant process improvement benefits are possible on account of

o Specialized functional skills (e.g. qualified design engineers)
o High process maturity (given the vendor's domain specialization).
l The functional outsourcing model is likely to see the highest BPM benefits, though less
cost arbitrage benefits.
l SECURITY THE KEY CONCERN.

SECURITY & IP PROTECTION

l All publicly traded corporations in the US, will be governed by the Sarbanes-Oxley Act.
The act mandates implementation of the CobIT security framework. Which in turn
requires:
l Authentication
o Access Control
o User account management
o Credential life-cycle management
o Non-repudiation
o Audit controls
l A documented SECURITY POLICY must be in place. And enforced.
l Corporations must consider Auto-ID technologies (RFID being the fore-runner). RFID
solutions, integrated with the enterprise Identity and Role-based Access Management
applications, successfully enable corporations to meet CobIT requirements, cost
effectively.


diginotes

THE VENDOR SIDE PERSPECTIVE

The discussion so far looks at the needs of the outsourcing corporations (the PRINCIPALS). Cl
early, VENDORS (the outsourcee), must be capable of delivering to these needs.

However, in addition, the vendors must manage their own business interests. Important vendor
-side issues are:
A. Project Portfolio Management (PPM)
B. Risk Management
C. Delivery Workflow Design
D. Resource Management
E. Process Improvement

A. PROJECT PORTFOLIO MANAGEMENT

A well defined PPM system, will lead to better financial performance and strategic alignment for
the vendor.Selection criteria for projects should be based on:
· NET PRESENT VALUE INDICATOR based on project cash flows.
· FEASIBILITY INDICATOR can the project be accepted given available capacity,
resources,skills and knowledge?
· STATEGIC INDICATOR what is the projects potential to create a sustainable longer
term competitive advantage on attributes such as market attractiveness, competitive
activity, organizational capability, risk assessment - internal and external.

By appropriately assigning values and weightings to each of the above, a PRIORITIZATION
INDEX can be arrived at. The factors that drive the PRIORITIZATION INDEX must be reviewed
and re-baselined periodically. For e.g. the project cash flows will change, capacities increased,
and so on.

The feasibility checks and project scheduling should particularly focus on loading of STRATEGIC
RESOURCES.

The PRIORITIZATION INDEX will then provide the rationale for inducting projects, closure of
“bleeding” projects, and resource conflicts.

B. RISK MANAGEMENT

As per Standish, only 26 % of all projects are considered successful. For IT projects, it is only
16%!

The top three reasons for unsuccessful IT projects are:
· Inadequately defined requirements
· Delays in change request management
· Constraints on critical chain resources.
With fixed-bid projects on the rise, vendors must de-risk themselves. In particular, they must pay
special attention to mitigate risks on the top three factors.

C. DELIVERY WORKFLOW DESIGN

The delivery workflow must be designed with the following considerations:

· THROUGHPUT MAXIMIZATION
Maximum value (EARNED VALUE) in the shortest elapsed time (CYCLE TIME).
There is a strong case for project managers to apply “CRITICAL CHAIN” concepts, espoused
by Eliyahu Goldratt. Proponents of the Critical Chain concept, claim 25 to 300% reduction in
total project cycle-times. Certainly there are improvement opportunities, even if they are not as
dramatic as the evangelists claim. Tough times require that no improvement opportunity be
squandered.


diginotes

· FOCUS Management must focus on a few GLOBALLY critical issues, rather than “policing”
every activity. Critical Chain project management makes a compelling argument for focusing
exclusively on the PROJECT BUFFER and the STRATEGIC RESOURCES.

· WORK-IN-PROGRESS MANAGEMENT’
o The list pending items in a persons INBOX, must as small as possible. At the same
time, no one should be idle.
o There should be, if possible zero latency, between actual start of work, and it's
assignment.
o Like in a “relay race”, succeeding task members must be primed up to receive the
baton from predecessors.

D. RESOURCE MANAGEMENT
· Golden rule: Strengthen the weakest link (on the critical chain) on the highest value chain (the
project with the highest prioritization index).
· The resource pool, particularly STRATEGIC RESOURCES must be accessible as GLOBAL
RESOURCE POOL. STRATEGIC RESOURCES are those that typically
· yield the highest value (billable rate high, assignable to projects of high prioritization
index)
· Are key to success of a project / activity. For a project activity type, attributes of a
strategic resource must be defined.
· Loading of strategic resources must be considered before acceptance of new projects
(on that type).
· The resourcing team must therefore focus on the strategic resource pool:
o To de-risk. Ensure sufficient back-ups and alternatives available.
o Training and resource development efforts are oriented towards converting non-
strategic resources to strategic.

E. PROCESS IMPROVEMENT

Process Centric IT Lifecycle is not new. It has been long since ISO 9000 certified companies
graduated to SW-CMM and to P-CMM and CMMi. All these have provided roadmap to an
organization for the long-term process improvement. Despite this, customer satisfaction and
eliminating defects remain a continuous challenge. Process maturity and quality will continue to
be the key selection criteria of “the vendor” by an outsourcing company. In addition, vendor will
have their own set of internal challenges to make best use of their resources/assets and comply
to strict financial compliances.

CMMi helped the organizations to identify processes and subsequently their capabilities matured
over the years however processes implemented with CMMi also has variability and most critically
it is important to understand which of these processes are not bringing any VALUE to the
organization or have become obsolete. In light of this, it is about time that IT organizations must
look into icing their CMMi processes with Six Sigma that will provide the data to understand the
variability of the processes and further improvement. Not only this, there has to be an end-to-end
integration between various standalone process frameworks like ITIL, CoBIT, SOX etc.


Demystifying Outsourcing and Global Services Delivery

More Related Content

What's hot (16)

Viewers also liked (7)

Similar to Demystifying Outsourcing and Global Services Delivery (20)

Recently uploaded (20)

Demystifying Outsourcing and Global Services Delivery