9 min read • Guide 827 of 877
Value Stream Optimization
Value flows when waste is removed. GitScrum helps teams visualize and measure their value stream, identifying bottlenecks and improvement opportunities.
Value Stream Basics
Understanding Flow
VALUE STREAM CONCEPT:
┌─────────────────────────────────────────────────────────────┐
│ │
│ VALUE STREAM = Idea → Production │
│ │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ ││
│ │ [Idea] → [Design] → [Develop] → [Test] → [Deploy] ││
│ │ ││
│ │ Each box = Step that adds value ││
│ │ Each arrow = Handoff (potential delay) ││
│ │ ││
│ └─────────────────────────────────────────────────────────┘│
│ │
│ ─────────────────────────────────────────────────────────── │
│ │
│ TWO TYPES OF TIME: │
│ ────────────────── │
│ │
│ PROCESS TIME: │
│ Actual time spent working (value-added) │
│ │
│ WAIT TIME: │
│ Time waiting (not adding value) │
│ │
│ ─────────────────────────────────────────────────────────── │
│ │
│ TYPICAL BREAKDOWN: │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ ││
│ │ LEAD TIME: 10 days total ││
│ │ ││
│ │ ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░████ ││
│ │ ◄─────────── Wait time: 9 days ───────────►◄ Work ► ││
│ │ 1 day ││
│ │ ││
│ │ PROCESS EFFICIENCY: 1/10 = 10% ││
│ │ ││
│ │ 90% of time = Waste (waiting) ││
│ │ 10% of time = Actually working ││
│ │ ││
│ └─────────────────────────────────────────────────────────┘│
│ │
│ GOAL: Increase process efficiency by reducing wait time │
└─────────────────────────────────────────────────────────────┘
Value Stream Mapping
Creating the Map
VALUE STREAM MAP:
┌─────────────────────────────────────────────────────────────┐
│ │
│ EXAMPLE: Feature Development Value Stream │
│ │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ ││
│ │ ┌────────┐ ┌────────┐ ┌────────┐ ┌────────┐ ││
│ │ │ GROOM │ │ SPRINT │ │ DEVELOP│ │ CODE │ ││
│ │ │ │──►│ PLAN │──►│ │──►│ REVIEW │ ││
│ │ └────────┘ └────────┘ └────────┘ └────────┘ ││
│ │ PT: 1h PT: 30m PT: 16h PT: 1h ││
│ │ WT: 5d WT: 0 WT: 0 WT: 1d ││
│ │ ││
│ │ ┌────────┐ ┌────────┐ ┌────────┐ ││
│ │ │ QA │ │ UAT │ │ DEPLOY │ ││
│ │ │ TEST │──►│ │──►│ │ ││
│ │ └────────┘ └────────┘ └────────┘ ││
│ │ PT: 2h PT: 30m PT: 15m ││
│ │ WT: 1d WT: 3d WT: 2d ││
│ │ ││
│ │ PT = Process Time (actually working) ││
│ │ WT = Wait Time (in queue, blocked) ││
│ │ ││
│ └─────────────────────────────────────────────────────────┘│
│ │
│ ─────────────────────────────────────────────────────────── │
│ │
│ TOTALS: │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ ││
│ │ TOTAL PROCESS TIME: ~21 hours (2.6 days) ││
│ │ TOTAL WAIT TIME: ~12 days ││
│ │ TOTAL LEAD TIME: ~14.6 days ││
│ │ ││
│ │ PROCESS EFFICIENCY: 2.6 / 14.6 = 18% ││
│ │ ││
│ │ 82% of the time, work is waiting, not being done! ││
│ │ ││
│ └─────────────────────────────────────────────────────────┘│
│ │
│ BIGGEST WAIT TIMES: Grooming wait (5d), UAT wait (3d) │
│ → These are the bottlenecks to address │
└─────────────────────────────────────────────────────────────┘
Identifying Waste
Types of Waste
SEVEN WASTES IN SOFTWARE:
┌─────────────────────────────────────────────────────────────┐
│ │
│ 1. WAITING: │
│ ─────────── │
│ • Waiting for requirements │
│ • Waiting for code review │
│ • Waiting for deployment window │
│ • Waiting for approval │
│ │
│ 2. OVERPRODUCTION: │
│ ────────────────── │
│ • Building features nobody uses │
│ • Over-engineering │
│ • Premature optimization │
│ │
│ 3. EXTRA PROCESSING: │
│ ──────────────────── │
│ • Unnecessary documentation │
│ • Bureaucratic processes │
│ • Re-entering data │
│ │
│ 4. MOTION: │
│ ────────── │
│ • Context switching │
│ • Tool switching │
│ • Finding information │
│ │
│ 5. DEFECTS: │
│ ─────────── │
│ • Bugs requiring rework │
│ • Misunderstood requirements │
│ • Production incidents │
│ │
│ 6. INVENTORY: │
│ ───────────── │
│ • Work in progress (WIP) │
│ • Unreleased features │
│ • Unmerged branches │
│ │
│ 7. HANDOFFS: │
│ ──────────── │
│ • Dev → QA → UAT → Ops │
│ • Each handoff = information loss + delay │
│ │
│ ─────────────────────────────────────────────────────────── │
│ │
│ IDENTIFY IN YOUR VALUE STREAM: │
│ Which wastes are causing your wait times? │
└─────────────────────────────────────────────────────────────┘
Optimization Strategies
Reducing Wait Time
OPTIMIZATION APPROACHES:
┌─────────────────────────────────────────────────────────────┐
│ │
│ BOTTLENECK: Code Review (1 day wait) │
│ ───────────────────────────────────── │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ ROOT CAUSE: ││
│ │ • Only senior devs review ││
│ │ • Large PRs take longer ││
│ │ • Reviews not prioritized ││
│ │ ││
│ │ SOLUTIONS: ││
│ │ • Train more reviewers ││
│ │ • Smaller PRs ││
│ │ • Review SLA (4 hour max wait) ││
│ │ • Pair programming (no review needed) ││
│ │ ││
│ │ RESULT: Wait time 1 day → 4 hours ││
│ └─────────────────────────────────────────────────────────┘│
│ │
│ ─────────────────────────────────────────────────────────── │
│ │
│ BOTTLENECK: UAT (3 day wait) │
│ ──────────────────────────── │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ ROOT CAUSE: ││
│ │ • Business users not available ││
│ │ • Batch UAT once per week ││
│ │ • No staging environment ││
│ │ ││
│ │ SOLUTIONS: ││
│ │ • Continuous UAT (flow, not batch) ││
│ │ • Feature flags (UAT in production) ││
│ │ • Automate acceptance tests ││
│ │ • Product owner validates continuously ││
│ │ ││
│ │ RESULT: Wait time 3 days → hours ││
│ └─────────────────────────────────────────────────────────┘│
│ │
│ ─────────────────────────────────────────────────────────── │
│ │
│ BOTTLENECK: Deployment (2 day wait) │
│ ─────────────────────────────────── │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ ROOT CAUSE: ││
│ │ • Manual deployment process ││
│ │ • Weekly deployment window ││
│ │ • Fear of breaking production ││
│ │ ││
│ │ SOLUTIONS: ││
│ │ • Automate deployment ││
│ │ • Continuous deployment ││
│ │ • Feature flags for safe rollout ││
│ │ • Automated rollback ││
│ │ ││
│ │ RESULT: Wait time 2 days → minutes ││
│ └─────────────────────────────────────────────────────────┘│
└─────────────────────────────────────────────────────────────┘
Measuring Improvement
Value Stream Metrics
VALUE STREAM METRICS:
┌─────────────────────────────────────────────────────────────┐
│ │
│ KEY METRICS: │
│ ──────────── │
│ │
│ LEAD TIME: │
│ Request → Production │
│ Measures total time to deliver value │
│ │
│ CYCLE TIME: │
│ Start work → Complete │
│ Measures team throughput │
│ │
│ PROCESS EFFICIENCY: │
│ Process Time / Lead Time │
│ Measures waste in the system │
│ │
│ ─────────────────────────────────────────────────────────── │
│ │
│ IMPROVEMENT TRACKING: │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ VALUE STREAM IMPROVEMENT REPORT ││
│ │ ││
│ │ METRIC BEFORE AFTER CHANGE ││
│ │ ────── ────── ───── ────── ││
│ │ Lead Time 14.6 d 5.2 d -64% ││
│ │ Cycle Time 2.6 d 2.1 d -19% ││
│ │ Efficiency 18% 40% +122% ││
│ │ ││
│ │ WAIT TIME BY STAGE: ││
│ │ ││
│ │ STAGE BEFORE AFTER IMPROVEMENT ││
│ │ ───── ────── ───── ─────────── ││
│ │ Grooming 5 d 1 d -80% ││
│ │ Code Review 1 d 4 hr -83% ││
│ │ QA 1 d 4 hr -83% ││
│ │ UAT 3 d 1 d -67% ││
│ │ Deploy 2 d 15 min -99% ││
│ │ ││
│ │ BIGGEST WIN: Deployment automation ││
│ └─────────────────────────────────────────────────────────┘│
│ │
│ VISUALIZE TRENDS: │
│ Plot lead time and efficiency over time │
│ Celebrate improvements, investigate regressions │
└─────────────────────────────────────────────────────────────┘
Continuous Improvement
Ongoing Optimization
VALUE STREAM KAIZEN:
┌─────────────────────────────────────────────────────────────┐
│ │
│ REGULAR REVIEW: │
│ ─────────────── │
│ │
│ Monthly: │
│ • Review value stream metrics │
│ • Identify current bottleneck │
│ • Plan improvement experiment │
│ │
│ Quarterly: │
│ • Full value stream mapping refresh │
│ • Update current state │
│ • Define future state │
│ │
│ ─────────────────────────────────────────────────────────── │
│ │
│ IMPROVEMENT CYCLE: │
│ ┌─────────────────────────────────────────────────────────┐│
│ │ ││
│ │ 1. MEASURE ──────────────► ││
│ │ ▲ │ ││
│ │ │ ▼ ││
│ │ │ 2. IDENTIFY ││
│ │ │ BOTTLENECK ││
│ │ │ │ ││
│ │ │ ▼ ││
│ │ 4. VERIFY ◄──────────── 3. IMPROVE ││
│ │ ││
│ │ 1. Measure current state ││
│ │ 2. Find the biggest bottleneck ││
│ │ 3. Implement improvement ││
│ │ 4. Verify it worked ││
│ │ (Repeat—the bottleneck will shift) ││
│ │ ││
│ └─────────────────────────────────────────────────────────┘│
│ │
│ ONE BOTTLENECK AT A TIME: │
│ ───────────────────────── │
│ Fixing a bottleneck moves it somewhere else │
│ That's good! It means you improved │
│ Now find the new bottleneck │
│ │
│ THEORY OF CONSTRAINTS: │
│ The bottleneck limits the whole system │
│ Improving non-bottlenecks doesn't help much │
│ Focus on the constraint │
└─────────────────────────────────────────────────────────────┘