Key Concepts
Software project management is about managing people, risk, and change — not just tracking tasks.
Managing People
The most important and most underestimated part of SE management. Motivation, trust, retention.
Project Planning
Breaking down work, estimating effort, scheduling, tracking progress. Plans will change — build in slack.
Risk Management
Identify risks → Assess likelihood/impact → Develop mitigation strategies → Monitor.
Project Metrics
Measure progress, velocity, quality. Use data to make decisions — not just gut feel.
Concept Deep Dives
Click each concept to expand — real examples, diagrams, pros & cons.
Managing People
When to Use
From day 1 as a tech lead or engineering manager.
Real-World Example
Google's Project Aristotle: psychological safety was #1 predictor of team performance — above skills or structure.
✓ Advantages
- High-performing teams outperform systems/processes
- Retention saves hiring costs
⚠ Watch Out
- No formula — each person is different
- Takes years to master
Project Planning
When to Use
Start of every project and every sprint. Planning ≠ prediction — it's a shared conversation.
Real-World Example
'Plans are useless, but planning is indispensable' — Eisenhower. The act of planning reveals unknowns.
✓ Advantages
- Shared team understanding
- Identifies risks early
- Basis for tracking progress
⚠ Watch Out
- Estimates are always wrong
- Over-planning = wasted effort
- Plans become outdated
Risk Management
When to Use
Start of every project. Revisited every sprint/milestone.
Real-World Example
Risk: 'Key developer may leave.' Mitigation: pair programming, documentation, bus factor reduction.
✓ Advantages
- Reduces surprises
- Forces proactive thinking
- Stakeholder communication
⚠ Watch Out
- Easy to identify risks, hard to mitigate them
- Risk registers become stale
- Likelihood: medium
- Impact: high
- Mitigate: pair-program, document
- Contingency: hire contractor
- Likelihood: high
- Impact: medium
- Mitigate: agile + change management
Project Metrics
When to Use
Throughout the project — measure what matters, not what's easy to measure.
Real-World Example
DORA metrics: Deployment Frequency, Lead Time, MTTR, Change Failure Rate — the 4 that actually predict performance.
✓ Advantages
- Objective basis for decisions
- Early warning signals
- Accountability
⚠ Watch Out
- Goodhart's Law: metric becomes target and stops being good measure
- Gaming the metric
Quick Reference
- 1Software project management: scheduling, risk management, people management.
- 2People management: most important skill — motivation, trust, team building.
- 3Planning: break down work, estimate effort, schedule, track. Plans will change.
- 4Risk management: identify, assess likelihood/impact, mitigate, monitor.
- 5Project metrics: velocity, quality, DORA metrics — data-driven management.
- 6Communication: regular status, clear escalation paths, stakeholder management.
From the Book & Beyond
Case Study — Alice, Dorothy & the Motivation Problem (Fig 22.8)
Sommerville's motivation case study (Figure 22.8) follows Alice, a project manager at an alarm-systems company entering the assistive technology market with a six-person team. Months in, Dorothy — the team's hardware expert — starts coming in late, her work quality drops, and she stops talking to teammates. The cause isn't laziness: the product direction left her writing C code instead of using her hardware interfacing skills, and she fears her career is stalling. Alice's fix is textbook people management — more design autonomy plus training that opens future opportunities. The takeaway: motivation problems are usually job-design problems in disguise.
2026 Perspective — Remote Teams, DORA Metrics & AI Agents
Managing engineers in 2026 means managing people you rarely see — and "teammates" that aren't people at all. Hybrid and remote-first teams are normal, so Sommerville's informal communication now flows through chat threads and recorded standups. Gut-feel progress tracking is giving way to DORA metrics — deployment frequency, lead time, change failure rate, time to restore — as the standard team health check. And the newest challenge: AI coding agents as a resource to be planned, reviewed, and budgeted, shifting the manager's question from "who writes this?" to "who verifies it?" Motivation and trust still decide everything else.
Quiz — Test Yourself
Think through your answer first, then reveal.