First, some interesting numbers:

1

First, some interesting numbers:

~2,000

682@3

~80

2

>200

[To be defined on March 23]

2

“Costs and Risks on Software Acquisition Projects”

Dr Maliha HaddadAssistant Professor

Management Science DepartmentSchool of Business and Public Management

George Washington UniversityWashington, DC

Dr. Anita J. La SalleProfessor and Chair, Information Technology Department

Director, Masters of IT Management ProgramKogod School of Business

American UniversityWashington, DC

From

SEPG 2

004

– M

arch

10.

3

Focus of Presentation:Outsourced Software Projects incur hidden costs and

substantial risks.

Role of Organizations re Software Projects

Organization is sole

developer

(In-House Projects)

Organization contracts for

software

(Out-House Projects)

Organization is prime,with one

or more external sub-contractors

Domain of interest

First,

4

Focus of Presentation:Outsourced Software Projects incur hidden costs and

substantial risks.

Presents some “twists” on the widely accepted advantages of outsourcing software projects:

Looks at some hidden costs of being a customer in a software contracting relationship.

Identifies some risks inherent in outsourcing software projects.

Serves as guideline to both software contractors and customers engaged in software contracts.

5

Outsourced Software Projects incur hidden costs and substantial risks.

Some Research Questions (background):

Is it possible to identify “hidden” costs of outsourcing software projects?

Is there a relationship between the size of the project and the hidden costs?

Is there a minimum baseline cost for a contracted software project independent of its size or complexity?

Where and when are the outsourcing costs incurred?

What are the risks that are incurred by outsourcing projects?

6

Focus of Presentation: Outsourced Software Projects incur hidden costs and substantial risks.

What we will cover:Research into software project outsourcingInitial research findingsFindings about outsourcing hidden costsFindings about outsourcing risksMore on risksConclusions

7

Research Strategy: Are there hidden costs?

1. Explore Estimation

Processes Used

by Contracting

Organizations

2. Identify and

Quantify

Sub-Contracting

Costs

3. Develop

“hidden-cost”

Estimation Model

•The survey/interviews

•The organizations

•The projects


8

Research into organizational practices

Survey/interviews of 26 orgs. about acquisition practices. (Fed. Govt., Telecomm.,DoD, Financial Inst.)

Contracts for $30K to $50MProjects: Business, Eng., AI/ES, Hybrids


9

Study focus: $ to contracting organization during SDLC

to acquire, manage, control, and support the software contract.

Time effort of personnel to support the contract.

Other resources needed for contract support.

Before, during and

after product

deployment


10

Some initial research observations: [lack of] Organizational awareness of models [lack of] Formal institutionalized software

acquisition plans or project tracking plans Traditional “order and wait” scenarios [lack of] Software requirements document

specificity


11

Some initial research observations (continued): [lack of] Acquisition project management processes for:

Contract management Configuration management for tracking Internal/external personnel oversight

processes Tracking progress against requirements and

costs Artifact inspections Risk identification and management Metrics gathering


12

Most organizations did not have formal methodologies in place for estimating the projected cost of their involvement in contracted projects, nor do they have processes in place for scheduling their resources that must be involved in projects.

Most organizations do not have formal procedures for collecting and archiving metrics about contracted projects, they do not classify internal costs as part of the project, nor do they include internal costs as part of any economic feasibility analysis for a contracted project.


13

Some research results:The hidden costs of contracting software is substantial – mean value is 190% of the contract.Linear relationship between hidden costs and project size (for the study sample):

M = 2.2 * KLOC + 52 (person months)


14

When and where are the hidden costs generated within the contracting

organization?


15

Distribution of “hidden” costs across SDLC


16

Distribution of “hidden” costs by resource types


OtherConsultantsQAUsersManagement

Me

an

60

50

40

30

20

10

0OtherConsultantsQAUsersManagement

Me

an

60

50

40

30

20

10

0

17

Implications of Findings:

Results of the study support the need for contracting organizations to embrace recommended and evolving methodologies (such as SA-CMM).

Few of the studied projects were housed in organizations having maturity levels beyond the Initial Level (only eight of the target organizations were even aware of their maturity level).

The availability of project data was more of an artifact of individual initiatives than organizational policy.


18

Implications of Findings (continued):

Organizations that fail to adopt structured software acquisition processes may have many of the same project risk profiles as organizations that fail to adopt structure software development processes [we will return to this issue].

A formal, institutionalized, software acquisition project management process model will require that the organization plan all aspects of the acquisition, manage software requirements, track project team and contractor team performance, manage the project’s cost and schedule baselines, evaluate the products and services, and successfully transition the software to its support organization.


19

Implications of Findings (continued):must be Risk management integrated into all aspects of the project, and the organization should provide the training required by personnel involved in the acquisition.


20

The Next Steps:

• Given the data elicited during the last phase of the research and the results of the study, can we devise a risk assessment/management model that will capture the project risks that are unique to the software outsourcing domain?

• If we are able to identify sub-contracting risks, what are some mitigation techniques that contracting organizations can put in place to manage risk?


21

Some next step issues:

•What are some outsourcing “Risk Categories” that may impact schedule, performance, cost?

•How can these risks be managed (i.e., oversight techniques)?

•How can organizations work with contractors to devise mitigation and contingency plans?

•What are the costs of managing subcontracting risks?


22

The second part:Risk Identification and Management

as aMajor Component of

Customer-Contractor relationships


23

Nature of risks in software contracting

Impacts of contracting risks on organizations

Sources of software contracting risks:

•Customer Internal Risks•Contractor Internal Risks•Risks at the Interfaces

•Customer Internal Risks•Contractor Internal Risks•Risks at the Interfaces


24

• Inaccurate estimates of effort (time, scope, $)• Personnel knowledge (software + acquisition)• User availability and involvement• Specification of customer requirements• Contract specificity (processes and interfaces)• Creeping requirements• Unanticipated coordination and oversight• … and more …

• Inaccurate estimates of effort (time, scope, $)• Personnel knowledge (software + acquisition)• User availability and involvement• Specification of customer requirements• Contract specificity (processes and interfaces)• Creeping requirements• Unanticipated coordination and oversight• … and more …

What are some internal risk sources for customers?


25

What are some contractor internal risks?

[Adherence to process models such as SEI’s CMMI is meant to reduce or eliminate most of contractor’s risks

– or provide guidelines for risk management plan.]

[Additional problems arise if the contractor is sub-contracting!]


26

• Mutually accepted ambiguous contract

• Ill-defined interfaces

• Multiple contacts

• Antagonistic interfaces

• Deficient inspections

• Loosely defined checkpoints

• Unavailable testing criteria, processes, data, benchmarks

• Mutually accepted ambiguous contract

• Ill-defined interfaces

• Multiple contacts

• Antagonistic interfaces

• Deficient inspections

• Loosely defined checkpoints

• Unavailable testing criteria, processes, data, benchmarks

What are some of the sources of risk at the interface between the customer and the software contractor?


27

• Shared repositories

• Configuration management of artifacts

• Risk management program

• Quality assurance program

• Expectations of re-use and maintainability

• Missed schedules

• Shared repositories

• Configuration management of artifacts

• Risk management program

• Quality assurance program

• Expectations of re-use and maintainability

• Missed schedules



28

• Costs of tools and management software

• Incompatible deployment and development infrastructures

• Security

• Transient personnel

• Unanticipated direct costs

• Costs of tools and management software

• Incompatible deployment and development infrastructures

• Security

• Transient personnel

• Unanticipated direct costs



29

• Incompatible processes and standards• Activity synchronization• “Gutless” oversight• Manager end-around play• Negotiating change• Test site requirements• Contract termination and litigation

• Incompatible processes and standards• Activity synchronization• “Gutless” oversight• Manager end-around play• Negotiating change• Test site requirements• Contract termination and litigation



30

Process is still “King/Queen” – especially in an outsourcing environment.

The cost of outsourcing a software project is not the cost of the contract.

Outsourcing software projects incurs “hidden costs” that are substantial.

Outsourcing software projects incurs substantial risks: within the client’s organization, within the contractor’s organization, and at the interfaces between client and contractor.

Process is still “King/Queen” – especially in an outsourcing environment.

The cost of outsourcing a software project is not the cost of the contract.

Outsourcing software projects incurs “hidden costs” that are substantial.

Outsourcing software projects incurs substantial risks: within the client’s organization, within the contractor’s organization, and at the interfaces between client and contractor.

Conclusions:


31

Contact information:


Dr Maliha [email protected]

Assistant ProfessorManagement Science

DepartmentSchool of Business and Public

Management George Washington University

Washington, DC

Dr. Anita J. La [email protected]

Professor and Chair, Information Technology Department

Director, Masters of IT Management Program

Kogod School of BusinessAmerican University

Washington, DC

Documents

First, some interesting numbers: