2. PRESALE & PROJECT MANAGEMENT

2. PRESALE & PROJECT MANAGEMENT

The Life Cycle of A Large System Integration Project The Life Cycle of A Large System Integration Project The Life Cycle of A Large System Integration Project The Life Cycle of A Large System Integration Project

PRESALEPRESALE

Business Plan

Customer Relationship

Solution and Technology

Process

BUSINESS PLAN (1)BUSINESS PLAN (1)

Creative product or penetrating existing market

FEASIBILITY STUDY Market analysis (marketing person or outsourcing)

government statisticsmarket (scale, competitor)potential customers and their valuesstrong points and week points (predominance)risk analysis

SHORT/LONG TERM OBJECTIVEinitial investment, mid-term investmentrevenue, cash flow and sales channeltechnologyproduct (series) planning resource (facility, engineer, marketing, sales)possible solutions, initial product effort and time to market

BUSINESS PLAN (2)BUSINESS PLAN (2)

11stst

yearyear22ndnd

yearyear33rdrd

yearyear44thth

yearyear55thth

yearyear

TimeTime

RevenueRevenueM US$M US$

1010

InvestmentInvestmentRevenueRevenue

99

88

77

66

55

44

33

22

11

1010thth

yearyear

CUSTOMER RELATIONSHIPCUSTOMER RELATIONSHIP

Top management’s customer visit to show support

Take customer to site visit

Build friendship

TECHNOLOGY AND SOLUTIONTECHNOLOGY AND SOLUTION

Experiences: past projects and technology/solution involved

Influence the Request For Proposal (RFP)

Organize seminars for the customers

Philosophy of technology/solution selection

PRESALE PROCESSPRESALE PROCESSObjectives:Objectives:• Establish solid customer relationship to build up trust• Introduce company’s successful history, experiences,

solutions and leading technology• Help customer to build up the knowledge• To know what the customers want and what they prefer • To know customer’s budget• To influence the RFP (Request For Proposal) as much as

possible

Participants:Participants:• Marketing and sales• Program manager, Sometimes business director, • Engineers (system engineer and software engineer)• Sometimes the business director, even president

Activities:Activities:• Seminars and presentations• Project sites visit• Sometimes home site visit• Leisure contact

BIDDING (1)BIDDING (1)Prequalification:Prequalification:To reduce the number of bidders, so only qualifiedvenders will participate the bid.• Business license• Size, revenue of the company• Similar experiences of comparable size project

in recent five years • Solution summary

Procedure:Procedure:1. Issuance of RFP (Request For Proposal).

Once issued, no customer contact any more.2. Purchase RFP3. Bid opening: Declare price for each vendor4. Review and evaluation: Couple of months5. Bid closing: Announce winner

Contents of RFPContents of RFP• International open bid, deadline is set• Two sections: business and technical

BIDDING (2)BIDDING (2)

leads the teamleads the business teamlead a technical team

pricing and signaturehardware and softwarecustomer relationshiplegal terms and consulting administrators tasks

ParticipantsParticipants

PM:PM:CM:CM:

Chief Engineer:Chief Engineer:Business Director & Business Director &

Accountant:Accountant:Engineers:Engineers:

Marketing and salesMarketing and salesLegal:Legal:

AdministratorsAdministrators::

Time:Time:

Approximately two-three months

BIDDING (3)BIDDING (3)

Proposal Includes:Proposal Includes:

7 copies of the following items, one copy with original signature on every page

BUSINESSBUSINESS• Legal certificate of the business• Last three years’ financial reports• Liability• Certificate of CMMI level• Matrix (line by line, yes/no)

TECHNICALTECHNICAL• Assumptions• Proposed solution, technology and

tools• Effort estimation• Management method, schedule• Solution for every subsystem and

its info from the vendor• Matrix (line by line, yes/no)

MATERIALSMATERIALS• Price brakes down to parts

BIDDING (4)BIDDING (4)

Contract Includes:Contract Includes:

2 copies with original signatures on every page

• Scope of services (RFP & Proposal)

• Price• Payment schedule• Hardware and software• Confidentiality• Rights on data• Warranty• Limitation of liability• Indemnity• No solicitation• Arbitration (disputes)• Jurisdiction (laws apply)

THE LIFECYCLETHE LIFECYCLE

Risk Management

Requirement Management

Presa

le PP

P

Plans

Requirem

ent Im

pleme

nt InstallationA

nd T

esting

Desig

n Acceptan

ce Deliver

y En

d

2 yr 2 mo 12 mo 3 mo2 mo 3 mo5 mo 1 mo

ACTUAL DURATION: Total of 28 months

Op

erationB

acku

p

M1Planning

M2CollectionAnalysis

M5Testing

M6Delivery

M4Implementation

M3Design

6 MILESTONES

Program Management

Participants: Business, Marketing, Sales, Program Manager, Contract Manager, Subcontractor Manager (s), Project Manager, Hardware/Software Engineers, Customers and End-users

Project Management

THE MANAGEMENT SPECTRUMTHE MANAGEMENT SPECTRUM

Four P’s:

PeopleProductProcessProject

KEY ELEMENTS TO SUCCESSKEY ELEMENTS TO SUCCESS

TECHNOLOGYTECHNOLOGY• Technology Technology • Domain knowledge Domain knowledge • ExperiencesExperiences

CUSTOMERCUSTOMER• Culture Culture • RelationshipRelationship

METHODOLOGYMETHODOLOGY• Right processRight process

MANAGEMENTMANAGEMENT• Requirement Requirement • Risk Risk • Schedule (milestones)Schedule (milestones)• CostCost

TECHNOLOGYTECHNOLOGY• Technology Technology • Domain knowledge Domain knowledge • ExperiencesExperiences

CUSTOMERCUSTOMER• Culture Culture • RelationshipRelationship

METHODOLOGYMETHODOLOGY• Right processRight process

MANAGEMENTMANAGEMENT• Requirement Requirement • Risk Risk • Schedule (milestones)Schedule (milestones)• CostCost

THE PEOPLE(1)THE PEOPLE(1)

Needs of Human Being

Life

shelter

safety

rich

achievement

reputation

success

prestige

THE PEOPLE (2)THE PEOPLE (2)

People management maturity model: recruiting, selection, performance management, training, compensation, career development, organization and work design, and team/culture development.

PM-CMM is a companion to the CMM model, which guides organizations in the creation of a mature software process.

The Taxonomy of players:• Senior managers• HR• Project managers• Practitioners• Customers• End-users

THE PEOPLE (3)THE PEOPLE (3)The Software Team

Team 1 Leader Team 2 Leader Team m Leader

Team Members Team Members Team Members

Three Management Styles [Mantei 81]

Democratic Decentralized (DD)

Controlled Decentralized (CD) Controlled Centralized (CC)

No permanent leader, vary by tasks

Defined leaders for tasks and subtasks

Defined leaders for tasks and subtasks

Decision and approach are made by consensus

Decision are made at group level, implementation at subgroups (Team)

Top level problems are managed by a team leader

Horizontal communication Vertical communication Vertical communication

Best for difficult problems;

High morale; job satisfaction;

Too much communication

Good for simple problems;

Better for high modularity;

More efficient

Good for simple problems;

Better for high modularity;

More efficient

Group

THE PEOPLE (4)THE PEOPLE (4)

Four Paradigm [Constantine 93]

Closed paradigmTraditional hierarchy, good for software products

Random paradigmLoosely structured, depends on individual initiative, heavy communication

Open paradigmStructure between Closed and Random, heavy communication

Synchronous paradigmRely on the natural compartmentalization of the task, little communication outside task

THE PEOPLE (5)THE PEOPLE (5)

To achieve a high performance team:• Team members must have trust in one another• Skill distribution must be appropriate to the problem• Mavericks may have to be excluded from the team

Factors in constructing a team:• The difficulty of the problem to be solved• The size of the resultant program in lines of code or function points• The time that the team will stay together (team lifetime)• The degree to which the problem can be modularized• The required quality of reliability of the system to be built• The rigidity of the delivery date• The degree of communication required for the project

THE PEOPLE (6)THE PEOPLE (6)Coordination and Communication:

Formal, impersonal approachesPlan, tech memo, milestone, schedule, and deliverables

Formal, interpersonal proceduresQuality assurance, status review, and code inspection

Informal, interpersonal proceduresGroup meeting for info dissemination and problem solving

Electronic communicationEmail, E-bulletin board, video conferences

Interpersonal networkingInformal discussion with people inside/outside team

2 3 4 5 6

6 5 4 3 2

Project control tools

Source codeRepository data

Public bulletins

Code inspection

Electronic mailStatus review

Discussion with peers

Group meeting

Req. reviews

Design reviewsDocumentsProject milestonesError tracking reports

Use of coordination technique

Valu

e of co

ord

inatio

n tech

niq

ue

THE PEOPLE (7)THE PEOPLE (7)

• A jelled team is a group of people so strongly knit that the whole is greater than the sum of the parts, the probability of success goes way up

• A jelled team is a group of people so strongly knit that the whole is greater than the sum of the parts, the probability of success goes way up

• It is difficult to find a challenging and interesting project, but not as difficult as finding a jelled team, in which your creativity, energy, and happiness can be maximized

• It is difficult to find a challenging and interesting project, but not as difficult as finding a jelled team, in which your creativity, energy, and happiness can be maximized

THE PRODUCT (1)THE PRODUCT (1)

The dilemma of software project manager at the beginning of a project: Quantitative estimations and an organized plan before solid information is available (before requirement collection/analysis)

Understand the overall characteristics of the productRefer to past projects with similar scale, technology, and functions Software scope at system level based on RFP and PROPOSAL, which must be unambiguous and understandable at the management and technical levels:

• context• information objectives• function and performance

THE PRODUCT (2)THE PRODUCT (2)

Problem decompositionExample: a new word-processing product with unique features: voice and keyboard input; automatic indexing and table of content; automatic copy edit; page layout capability, etc.

Input: voice learningvoice recognitionkeyboard input

Automatic copy edit: spell checking, sentence grammar checking reference checkingsection and chapter reference validation

…

THE PROCESS (1)THE PROCESS (1)The software development process models:

select right process model that is best fit the project for the team

• The Waterfall (linear sequential) model• The Spiral (prototyping) model• The Iterative (incremental) model

• The RAD model• The WINWIN spiral model• The component-based development model• The concurrent model• The formal method model• The fourth generation technique model

• CMMI (Capacity Maturity Method Integration)• SEI (Software Engineer Institute)

THE PROCESS (2)THE PROCESS (2)

Model selection based on which process model is most appropriate for:

1. The characteristics of the project2. The customers and parishioners3. The project working environment

Common framework activities:1. Customer communication2. Planning3. Risk analysis4. Engineering 5. Construction and release6. Customer evaluation

THE PROJECT (1)THE PROJECT (1)

Signs that indicate that indicate that a project is in jeopardy:

PEOPLE REQUIREMENT RISK FINACE

• None technical split in the team• Software people do not understand their customer’s needs• The project scope is poorly defined• Changes are managed poorly• The chosen technology changes• Business needs change (or are ill-defined)• Financial difficulties• Deadline is unrealistic• Users are resistant• Sponsorship is lost• Lack of skill sets in the team• Avoid best practices and lessons learned

THE PROJECT (2)THE PROJECT (2)

Five-part commonsense approach [Reel 99]• Start on the right foot• Maintain momentum• Track progress• Make smart decision• Conduct a postmortem analysis

To manage a successful project is to manage problems, i.e.,

• to avoid problems• to reduce the degree of difficulties• to have a plan/solution before the problems occur.

PROJECT CONSTITUTIONPROJECT CONSTITUTION

Origin

Definition

Initiator

Manager

Team

Resource

Constitution

TYPICAL CONTENTS:

Project formal nameProgram manager and contactProject manager and contactProject target and deliverableProject time tableProject resource, budget, vendor

FACTS OF PMFACTS OF PM1995 vs 1998• The cost of failed projects went down from $81billion to $75 billion• Decrease in cost overruns from $59 billion to $22 billion

In 1998• 26% of information technology projects succeed in meeting scope, tim

e, and cost goals• 46 percent of IT projects completed over budget and past deadline• 28% failed

2001 vs 1995• Time overruns significantly decrease to 63%, compared to 222%• Cost overruns were down to 45%, compared to 189%• Required features and functions were up to 67%, compared to 61%• 78000 US projects were successful, compared to 28000• 28% of IT projects succeeded, compared to 16%

The Standish Group, “1998 CHAOS Report” & “CHAOS 2001”