3. Project Planning and Cost Estimation

3.1 What Is a Project?

Task: a fragmentary piece of work with few time or budget constraints

Project: planned work that must achieve a specific goal within limited time and cost

• It is completed by a project team led by a project manager (PM) and needs resources.

Work: a single unit of work performed inside a project

• The first three are the same and only the last one differs: a production system has Method, production resources have Money

Essentials for a successful SW project: people, resources (lab equipment, materials), capital, technology, time

What goes into a SW project plan: work plan, work method (technology) plan, resource plan, organization plan, schedule, budget plan, risk management plan, maintenance plan

3.2 Project Identification

System Request: a document explaining why the system should be built and what value it creates

• Components: project sponsor, business need, business requirements, business value, special issues

Project sponsor: a stakeholder

Business requirements: why the system is built + an outline of the functions it gives the organization

Work plan: provides the WBS (work breakdown structure)

Feasibility analysis

• Components: economic feasibility, technical feasibility, organizational feasibility, legal and environmental feasibility, schedule feasibility

3.3 Project Management and the Project Manager

Project management life cycle: initiation → planning → execution → maintenance

Project management: managing a project to reach its goals under the given constraints

• Components: work planning, risk assessment, resource estimation, organization, resource acquisition, work direction, progress tracking, result analysis

Main roles of the project manager (PM)

1. Planning, winning the contract, preparation

• Contract support, planning, writing the proposal, writing the WBS, assigning work, securing resources, cost estimation, scheduling, risk management

2. Management

• Managing overall progress: design, construction, testing, approval, etc.
• Budget management, progress management, keeping the work on course, performance reporting

3. Maintenance

• Result analysis, acceptance support, maintenance support, sales support for finding new projects

3.4 Work Planning and Scheduling

Typical management tools: WBS, CPM network, Gantt chart, PERT chart

WBS: the project plan broken down into levels

CPM network: represented by a PERT chart.

• It visualizes well but can miss some factors.

PERT chart: a graphical representation of the timeline of individual tasks

1. Components

• Node: event or milestone (circle or rectangle)
• Vector: lines showing tasks and the dependencies (precedence) between them

2. How to draw it
   1. List the subtasks
   2. Draw the CPM network diagram
      • Find every possible path in the CPM network and find the critical path (the longest path)

Gantt chart: a bar chart showing the project schedule

Schedule: a list of the activities, deliverables and milestones in a project

• Components: start and end dates, resources and similar information
• Process: plan schedule management → define activities → sequence activities → estimate resources → estimate durations → develop the schedule
• Types
  1. Master: a simple task list with a timeline
  2. Milestone: summary tasks
  3. Detailed: tracks every activity, used in large projects
• 7 principles: gather stakeholder input, refer to past projects, account for risks, account for non-working time, define the critical path, record assumptions, include milestones

3.5 Software Development Cost Factors

Research and planning, technology used, project size, UI/UX requirements, staff employment model, scale of maintenance and support

3.6 Types of Project Cost Estimation Models and the Procedure

Cost estimation models

1. Heuristic techniques
   • COCOMO (Constructive Cost Model): the representative heuristic technique, a LOC-based cost estimation regression model
     • LOC (Line of Code) = counting how many lines the program has
2. Analytical estimation techniques
   • Function Point Analysis (FPA): computes cost from the number and complexity of functions
   • Putnam model: a parametric model that estimates effort, time and defects
   • Price-To-Win estimation model: based on market conditions and competitors' prices
   • Machine learning based models
   • IFPUG (International Function Point Users Group): the function point measurement standard provided by the International Function Point Users Group

5 steps to prepare a cost estimate

Big picture (module map) → what to build it with (technology) → what to build (SRS) → how it looks (UI) → split the work and price it (WBS)

1. Write the project overview and a high-level module map

• Use cases, classes, methods, DB, UI, infrastructure and so on all go in here

2. Research and choose the required technology
3. Draft the SRS (software requirements specification)

• Requirement set, constraints, target users and so on

4. Detailed UI design
5. Write the WBS

3.7 LOC-Based Project Cost Estimation

• Basic COCOMO: reflects only a few constants (effort and schedule are the key variables)
  • Has estimation methods for organic, semi-detached and embedded systems
• Intermediate COCOMO: additionally reflects 15 software engineering cost attributes
• Detailed COCOMO: splits the whole software into several modules and estimates and applies effort multipliers

3.8 Function Point-Based Project Cost Estimation

Function Point (FP): a unit that measures the amount of functionality a system provides to users

• How many functions there are and how complex they are
• Industry standard

5 components

• ILF vs EIF: told apart by who defined it
• One search = EQ (inquiry) + EO (output)

Supporting terms

5 steps of function point estimation

1. Identify the function point types
2. Rate the complexity (simple / average / complex)
3. Compute the unadjusted function points (UFP)
4. Apply the total degree of influence of system characteristics (TDI)
5. Compute the final function points

3.9 Organization Planning

Ways to structure a team: hierarchical, chief programmer, matrix, democratic

Hierarchical team: tree structure, a sub-team leader for each functional sub-team

• Lower levels have detailed knowledge, upper levels understand the whole broadly
• Suited to large projects

Chief programmer team: centered on the chief programmer

• Chief: planning, specification, design (up to implementation)
• Backup: technical peer
• Secretary: supports administrative work
• Specialists: do the implementation

Matrix team: people belong to both a department (technology) and a project

• Project teams disband when the project ends, so skills don't accumulate → this solves that
• US aerospace industry in the 1950s; in Korea, LG

Democratic team: everyone discusses one-on-one

• The leader decides only when there is no consensus

Typical SW development team: PM, business analyst, SW architect, developers, QA, etc.

• The development team (offshore) communicates with the client (on-site)

Agile development team: typical team + PM, UX/UI designer, QA engineer

3.10 Risk Management Planning

Risk management plan: a plan to find risks in advance, meet deadlines, communicate, and allocate money to eliminating risks

Technical risk: risk caused by technical problems or faulty tools

• Factors: poor code quality, constantly changing requirements, wrong technology choices, integration problems
• Mitigation: get all existing documents and discuss problems with the previous team. New technology makes the risk bigger

Operational risk: risk that arises in the process when development goals are unclear

• Factors: unreasonable or inaccurate deadlines, wrong time estimates, lower productivity (delays, staff burnout), poor management, scope changes (scope creep), no communication with stakeholders, communication mistakes
• Rushing by skipping important stages, lack of skilled developers, and unstable workloads are also risks

Business risk: risk that arises from the company's business interests and operations

• Factors: unmet stakeholder expectations, undefined metrics, inconsistent priorities, budget problems, low engagement, no partnership agreement, unpredictable external risks
• No partnership agreement → solved with a written contract
• External risks: changes in law, government regulations, economic changes, natural disasters, etc. → minimized with insurance and staying up to date

3.11 Business Feasibility Analysis

Feasibility analysis: reviewing and evaluating every aspect of a project

• Types: economic, technical, organizational, legal and environmental, schedule

Economic feasibility: whether it makes money

• Criteria: present value (PV), net present value (NPV), return on investment (ROI), break-even point (BEP)
• PV: future money converted to its value today
• NPV: present value of money coming in - present value of money going out
• ROI: return relative to investment
• BEP: revenue = total cost, the point with neither loss nor profit

Technical feasibility: whether it can be built technically

• The organization's technical skill, availability of existing technology, ability to meet HW and SW requirements, ability to secure resources and technology, maintainability

Organizational feasibility: whether the organization can support it

• The sponsor's willingness to invest, managers' understanding and budget, user level, staffing (in-house and external), employees' willingness to accept it and their skills, ability to handle new technology, job security

Legal feasibility: whether there are no legal restrictions and ethical standards are met

• Contracts, data protection law, medical law, certificates, licenses, copyright, local ordinances, etc.

Schedule feasibility: whether it can be finished before the deadline

• Missing the deadline can mean the project fails its purpose

3.12 Project Management Standards

Standards: rules everyone on the project follows

• Types: document templates, work procedures, software requirements specification (SRS), software design description (SDD), UI design, coding

• SRS → information needed for the design phase, SDD → information needed for the implementation phase