outsourcing - a fact based decision? - ifpug proceedings/isma1-2006... · carl-henric svanberg, ceo...
TRANSCRIPT
Outsourcing – A Fact Based Decision?
ISMA Conference, September 2006, San Diego
Pierre Almén & Frank [email protected] & [email protected]
www.compassmc.com
22006-09-15
Agenda
Who are we?
Background
Fact Based Decision – Why
Fact Based Decision – How
Fact Based Decision – Summary
Outsourcing Vendor Evaluation – Case Study
42006-09-15
Compass - A Global CompanyPrivately owned, independent global consultancy
Founded in 1980, 18 offices in 15 countries
Over 450 active clients in 37 countries
Over 1500 client operations analysed in detail in 2005
Offices in:AustraliaAustriaBrazilCanada FinlandFranceGermanyItalyNetherlandsSingaporeSouth AfricaSpainSwedenUKUSA Country of operation
62006-09-15
Different IT views
CEO VIEWCIO VIEWIT Unit IT Total IT BusinessCost Volume Cost Value
Desktop -32% +283% +154% ?
Mainframe -66% +297% +5% ?
Network -42% +210% +80% ?
% Change 5 years
What is the business demand driving IT volumes ?Are IT and business operations aligned to maximise business impact?
72006-09-15
Market
IT services in Europe (source: Forrester Research)
10% yearly growth 2006-2011 for Application outsourcing, IT in total 6%
Outsourcing plans 2006: Applications 53%
Engineer examination
USA: 400.000 / year
India: 375.000 / year (and growing)
China: Today 5 times more engineers compared to USA (and rapidly growing)
India Sourcing MarketServices and Sourcing market: 17.2 billion $ April 2004 – March 2005 (source: National Association of Software and Service Companies, New Dehli)
Doubling every 3 years
65% of world AD/AM offshore and 46% of BPO (call centers, HR, ticket booking etc) (source McKinsey)
2% of BNP year 2000, now 5%, 7% year 2010 (source Nasscom/McKinsey)
82006-09-15
Outsourcing FailuresCauses
(May 2004 study ”Leading Causes of Outsourcing Failures”)
Buyer’s unclear expectations up front as to its objectives
Parties interests are aligned up front but become misaligned as business environment / needs change
Provider’s poor performance against SLA
Carl-Henric Svanberg, CEO Ericsson (source: DI June 23 2005)
“Many companies outsource on incorrect grounds”
“If the business is not efficient, the solution is not to outsource. You must improve your business first. Then a sourcing decision could be considered”.
102006-09-15
Why Software Measurement Has Value
Companies that don’t:
On-time projects: 45%
Late projects: 40%
Cancelled projects: 15%
Defect removal: Unknown
Cost estimates: Optimistic
User Satisfaction: Low
Software status Low
Staff morale: Low
Companies that measure:
On-time projects: 75%
Late projects: 20%
Cancelled projects: 5%
Defect removal: >95%
Cost estimates: Accurate
User Satisfaction: High
Software status High
Staff morale: High
Source: Capers Jones IFPUG 2001
112006-09-15
The Compass Effect
Over a 4 year period, regular clients are on averagemore than 36% more efficient than first time clients
132006-09-15
Selection of projects / applications (1)
Make an inventory of all applications / projects and analyze based on Strategic value and Operational Importance
If a specific application fails, how soon will your CEO find out?At what level is this application sponsored? Does application investment and mix apply to current business strategy?
Operational ImportanceHigh
High
Low
Low
StrategicValue
Strategic, High Impact
20%
Strategic, Low Impact
14%Non-strategic,
High Impact12%
Non-strategic,
Low Impact54%
142006-09-15
Selection of projects / applications (2)
Primary selection criteria are Strategic value and Operational importance
Other selection criteria that can be used:
Skill constraintsBusinessTechnical
Phase out ofPlatformProgramming language etc
Flexibility
152006-09-15
Expandable Metrics ScopeOverview
Time Reporting - Financial Reporting - Quality ReportingGenerally available information
Time Reporting - Financial Reporting - Quality ReportingGenerally available information
DevelopmentAll or specific projects
DevelopmentAll or specific projects
MaintenanceAll or specific applications
MaintenanceAll or specific applications
Project/App.#1
FP Level 3
Project/App.#1
FP Level 3
Project/App.#2
FP Level 4
Project/App.#2
FP Level 4
Project/App.#3
FP Level 4
Project/App.#3
FP Level 4
Project/App.#4
FP Level 5
Project/App.#4
FP Level 5
Project/App.#5
FP Level 5
Project/App.#5
FP Level 5
Project/App.#…
FP Level 6
Project/App.#…
FP Level 6
Starting with application development and maintenance basics and expanding upwards
162006-09-15
Expandable Metrics ScopeBasic Building Block
Sample performance indicators;Cost per chargeable hour Cost per worked hour
EmployeesContractors
Resource utilization Cost distributionResource distribution (employees/contractors)Contractor usage Training & EducationManagement & AdminUser, Customer and Employee Satisfaction
Time Reporting - Financial Reporting - Quality ReportingGenerally available information
Time Reporting - Financial Reporting - Quality ReportingGenerally available information
DevelopmentAll or specific projects
DevelopmentAll or specific projects
MaintenanceAll or specific applications
MaintenanceAll or specific applications
Project/App.#1
FP Level 3
Project/App.#1
FP Level 3
Project/App.#2
FP Level 4
Project/App.#2
FP Level 4
Project/App.#3
FP Level 4
Project/App.#3
FP Level 4
Project/App.#4
FP Level 5
Project/App.#4
FP Level 5
Project/App.#5
FP Level 5
Project/App.#5
FP Level 5
Project/App.#…
FP Level 6
Project/App.#…
FP Level 6
• Based on available time, financial and quality reporting and interviews with management
• Covers application development, maintenance and management & administration
• Information period can be e.g.latest 6 or 12 months
172006-09-15
High Level Cost EfficiencyCost per FTE / per Billable Hour
0
20
40
60
80
100
120
140
160
Total annual cost perFTE (K$)
Cost per chargeablehour ($)
Cost comparison
XYZ 2004XYZ 2005Ref Mean 2005
182006-09-15
Expandable Metrics ScopeDevelopment and Maintenance
Sample performance indicators;Cost efficiency within application maintenance Project durations, delivery precision Operation and development environments User/Customer involvementResource split by phases Resource estimationsChange and Release management Scope Creep (hours)
Time Reporting - Financial Reporting - Quality ReportingGenerally available information
Time Reporting - Financial Reporting - Quality ReportingGenerally available information
DevelopmentAll or specific projects
DevelopmentAll or specific projects
MaintenanceAll or specific applications
MaintenanceAll or specific applications
Project/App.#1
FP Level 3
Project/App.#1
FP Level 3
Project/App.#2
FP Level 4
Project/App.#2
FP Level 4
Project/App.#3
FP Level 4
Project/App.#3
FP Level 4
Project/App.#4
FP Level 5
Project/App.#4
FP Level 5
Project/App.#5
FP Level 5
Project/App.#5
FP Level 5
Project/App.#…
FP Level 6
Project/App.#…
FP Level 6
• Based on previously collected information extended with available project / application information and interviews with project and system managers
• Initially focus on a sub-set of projects and a sub-set of applications
• Information period can be e.g.latest 6 or 12 months
192006-09-15
Cost EfficiencyApplication Maintenance
0
500
1000
1500
2000
2500
3000
$
XYZ 2004 XYZ 2005 Ref Mean 2005
Repair cost per problem
202006-09-15
Expandable ScopeProject and Application Size (Complexity)
Sample performance indicators;Development and maintenance productivityCost efficiency within development and maintenanceApplication quality Delivery CapacityScope Creep (functionality)Code re-use
Time Reporting - Financial Reporting - Quality ReportingGenerally available information
Time Reporting - Financial Reporting - Quality ReportingGenerally available information
DevelopmentAll or specific projects
DevelopmentAll or specific projects
MaintenanceAll or specific applications
MaintenanceAll or specific applications
Project/App.#1
FP Level 3
Project/App.#1
FP Level 3
Project/App.#2
FP Level 4
Project/App.#2
FP Level 4
Project/App.#3
FP Level 4
Project/App.#3
FP Level 4
Project/App.#4
FP Level 5
Project/App.#4
FP Level 5
Project/App.#5
FP Level 5
Project/App.#5
FP Level 5
Project/App.#…
FP Level 6
Project/App.#…
FP Level 6
• Based on previously collected information and extended with Function Points for selected projects/applications
• FP information initially only for selected sub-set of projects / applications and within accuracy level 3-6
212006-09-15
Development Productivity
0
2
4
6
8
10
12
14
Environment A Environment B
Development Productivity - FP per PM
XYZ 2004XYZ 2005Ref Mean 2005
232006-09-15
Critical Success FactorsKnow YOUR performance – Get a baseline
MEASUREMENT is a key component of outsourcing components
Remember that there will be RETAINED services
No matter how good a deal you cut with a supplier – if that deal does not meet the needs of the wider business it is NOT a good deal
Key to your creation of the RFP is keeping the business agenda at the top of YOUR agenda
Stakeholder analysis, peer review and business sponsorship are all key to a successful outsourcing
252006-09-15
Overview of the EngagementClient wanted to migrate a large suite of applications from Mainframe to Client/Server and achieve cost reductions
This involved a two-phase procurementApplication re-development, and ongoing maintenance and support
Infrastructure Services (hosting)
Client was anxious to learn about leading Indian outsourcers andto include them in the process
This included visits to six leading vendors’ off-shore facilities
262006-09-15
As expected, Indian vendors place a very large emphasis on process.
Strong process orientationSEI CMM Level 5, CMMI, ISO 9000, MBNQA, BS 7799, etc.Process maturity is used as marketing leverageBut since all vendors have the same processes, there is little real competitive advantage
Sound processes are essential to the off-shore business modelIn other words, when you have an analyst and programmer working 10,000 miles apart, you’d better give them some good processes!
A corollary: You can use off-shore outsourcing to “buy” process maturity –BUT – only to the extent that your retained organization can assimilate it.
272006-09-15
All vendors use similar staffing models. 30% on-shore and 70% off-shore is the typical resource split.
On-shore resourcesRequirements gatheringAcceptance testing and installationCritical supportOther client-facing functions
Off-shore resourcesTraditional analysis/design/code and testingSoftware maintenance
On-shore resources are often higher during requirements phase, and are typically lower (~20%) for pure maintenance mode
282006-09-15
All Indian vendors emphasize Human Resources and Knowledge Management functions.
Human ResourcesEach meeting included a high-ranking HR manager
Very good recruiting and hiring practices, extensive testing and a <1% offer rate
Extremely friendly work environments (“campuses”)
Emphasis on training – soft skills as well as technical
Knowledge ManagementMore important to off-shore work because of time and geographical considerations
All vendors stressed KM practices and tools
292006-09-15
Overall, the evaluation team members were quite favorably impressed with the Indian vendors.
Team member rankings, at the end of the off-shore visits, were almost identical
However, there was also agreement that even the lowest-ranked vendor could do a very good job with the scope of work
And, as it turned out, the vendor ultimately selected was in themiddle of the pack based on the off-shore site visits
Remember, this was only one step of the process!
302006-09-15
So, does this mean you should outsource everything you do to India? Probably not!
Things that work well off-shore:Tightly-bounded development projects, with good requirements
Routine application maintenance work
Things that DON’T work well off-shore:Projects with poor or ill-defined specs
Staff augmentation (using off-shore resources as contractors)
Critical-application maintenance (where on-site support and immediate response may be required)
“Around the clock” development
The bottom line is, you still have to use your judgment!
312006-09-15
The procurement process should be structured with many vendor interactions and evaluations.
Sample procurement lifecycle:Market and vendor researchRFI developmentRFI issuance and evaluation of vendor responsesRFP development and issuanceVendor question-and-answer sessionOff-shore site visitsOn-shore site visitsMid-term vendor presentationsReference checksVendor proposal evaluationVendor selection, due diligence and contract negotiations
322006-09-15
And a comprehensive set of criteria should be used to evaluate the vendor interactions.
Evaluation Criteria Weight Score Weighted ScorePeopleProcessesToolsSolution DesignAbility to DeliverFinancial PerformanceMeasurement and Continuous ImprovementLeadershipExperienceValue AddAbility to Work TogetherTransition
Sample Vendor Evaluation Worksheet
332006-09-15
Vendor Evaluation - SummaryDo your homework first – Develop a Fact-Based Sourcing Strategy
Be sure to at least consider the leading off-shore outsourcers, as they all have good people and impressive delivery capabilities
Use your judgment, some things don’t adapt well to off-shoring
And use multiple vendor interactions and evaluation criteria to ensure you pick the best vendor for you