Velocity Tracking and Improvement | Planning Guide
Measure and improve team velocity for predictable planning. GitScrum tracks sprint trends, commitment vs completed, and sustainable improvement patterns.
7 min read
Velocity measures how much work a team completes per sprint. It's a planning tool, not a performance metric. Understanding velocity helps teams commit to achievable work and stakeholders set realistic expectations. Improving velocity means removing friction, not working harder.
Understanding Velocity
What Velocity Is
VELOCITY FUNDAMENTALS
βββββββββββββββββββββ
DEFINITION:
βββββββββββββββββββββββββββββββββββββ
Velocity = Story Points completed per sprint
Example:
βββ Sprint 1: 34 points
βββ Sprint 2: 28 points
βββ Sprint 3: 32 points
βββ Sprint 4: 30 points
βββ Average velocity: 31 points
WHAT VELOCITY IS:
βββββββββββββββββββββββββββββββββββββ
β Planning tool ("Can we commit to this?")
β Capacity indicator ("How much can we do?")
β Trend tracker ("Are we improving?")
β Team-specific measure
β Relative measure
WHAT VELOCITY IS NOT:
βββββββββββββββββββββββββββββββββββββ
β Performance metric
β Productivity measure
β Comparable between teams
β Absolute measure
β Target to maximize
β Individual metric
VELOCITY RANGE:
βββββββββββββββββββββββββββββββββββββ
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β Team Alpha Velocity β
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ€
β β
β 45 β€ β
β β β High: 38 β
β 35 β€ β βββββββββββββ Average: 31 β
β β β β β β
β 25 β€ β β β Low: 24 β
β β β
β 15 β€ β
β ββββββββββββββββββββββββββββββββββββββββββββββββββββ
β S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 β
β β
β Average: 31 pts | Range: 24-38 | Variance: Β±22% β
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
Use RANGE for planning, not just average.
Expect varianceβit's normal.
Using Velocity
VELOCITY FOR PLANNING
βββββββββββββββββββββ
SPRINT COMMITMENT:
βββββββββββββββββββββββββββββββββββββ
Step 1: Know your velocity
βββ Last 3 sprints: 32, 28, 34
βββ Average: 31 points
βββ Range: 28-34
βββ Commit: 28-31 points (conservative)
Step 2: Account for factors
βββ Holiday week? Reduce 20%
βββ Team member out? Reduce proportionally
βββ New team members? Reduce for onboarding
βββ Adjusted: 25 points
Step 3: Pull from backlog
βββ Pull highest priority items
βββ Until you hit capacity
βββ Leave buffer for surprises
βββ Don't exceed adjusted velocity
LONG-RANGE PLANNING:
βββββββββββββββββββββββββββββββββββββ
Release planning example:
βββ Remaining backlog: 180 points
βββ Average velocity: 30 points
βββ Sprints needed: 180 Γ· 30 = 6 sprints
βββ Buffer (10%): +1 sprint
βββ Estimate: 7 sprints (14 weeks)
βββ Communicate as range, not promise
Present as:
"Based on current velocity, we estimate
6-8 sprints. We'll update as we learn more."
Tracking Velocity
Measurement
VELOCITY TRACKING
βββββββββββββββββ
WHAT COUNTS:
βββββββββββββββββββββββββββββββββββββ
Count as complete:
βββ Meets Definition of Done
βββ Potentially shippable
βββ Accepted by PO
βββ Full story points
Don't count:
βββ Partially done (0 points, not fraction)
βββ "Almost done"
βββ Tested but not deployed (if deploy = done)
βββ Carried over to next sprint
βββ Scope added mid-sprint
TRACKING IN GITSCRUM:
βββββββββββββββββββββββββββββββββββββ
Sprint β Close Sprint β Velocity recorded
Dashboard shows:
βββ Velocity per sprint
βββ Trend over time
βββ Average velocity
βββ Variance
βββ Commitment vs. completed
VELOCITY TREND:
βββββββββββββββββββββββββββββββββββββ
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β Velocity Trend (12 sprints) β
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ€
β β
β 40 β€ β β β
β β β β β
β 30 β€ β β β β β β
β β β β β
β 20 β€ β β β
β β β
β 10 β€ β
β ββββββββββββββββββββββββββββββββββββββββββββββββββββ
β Formation β Stabilizing β Improving β β
β β
β β Trend: Improving (+2 pts/sprint) β
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
Early sprints: Low (learning)
Middle: Stabilizing
Later: May improve as team matures
Improving Velocity
Legitimate Improvements
HOW TO IMPROVE VELOCITY
βββββββββββββββββββββββ
SUSTAINABLE IMPROVEMENT:
βββββββββββββββββββββββββββββββββββββ
Velocity improves through:
1. REMOVE BLOCKERS
βββ Faster code review turnaround
βββ Quicker stakeholder decisions
βββ Better tooling
βββ Clearer requirements
βββ Visible in retros
2. REDUCE WASTE
βββ Fewer unnecessary meetings
βββ Less context switching
βββ Reduced rework
βββ Smaller batch sizes
βββ Better definition of ready
3. IMPROVE PROCESS
βββ Better testing practices
βββ CI/CD improvements
βββ More effective ceremonies
βββ Clearer Definition of Done
βββ Collaborative refinement
4. REDUCE TECHNICAL DEBT
βββ Easier codebase = faster work
βββ Fewer bugs = less rework
βββ Better architecture = faster features
βββ Investment pays off over time
5. TEAM GROWTH
βββ More experience
βββ Better collaboration
βββ Skills development
βββ Domain knowledge
βββ Natural over time
NOT SUSTAINABLE:
βββββββββββββββββββββββββββββββββββββ
β Working overtime
β Cutting quality
β Skipping tests
β Reducing estimates to "get more done"
β Pressure to hit targets
Finding Improvement Opportunities
VELOCITY IMPROVEMENT ANALYSIS
βββββββββββββββββββββββββββββ
USE RETROSPECTIVES:
βββββββββββββββββββββββββββββββββββββ
Ask:
"What slowed us down this sprint?"
"Where did we wait?"
"What would we do differently?"
Common answers that improve velocity:
βββ "Waiting for code review"
β β Implement review SLA
β
βββ "Requirements were unclear"
β β Improve refinement
β
βββ "Deployment took forever"
β β Invest in CI/CD
β
βββ "Meetings fragmented my day"
β β Meeting-free blocks
β
βββ "Bugs in legacy code"
β Address technical debt
USE METRICS:
βββββββββββββββββββββββββββββββββββββ
Cycle time by stage:
βββ Development: 2 days (expected)
βββ Code Review: 3 days (too long!)
βββ Testing: 1 day (expected)
βββ Deploy: 2 days (too long!)
βββ Total: 8 days
Target improvements where wait is longest.
WIP ANALYSIS:
βββββββββββββββββββββββββββββββββββββ
High WIP = Low throughput
βββ Too much in progress
βββ Context switching overhead
βββ Nothing finishing
βββ Reduce WIP limits
Velocity Antipatterns
What to Avoid
VELOCITY ANTIPATTERNS
βββββββββββββββββββββ
USING VELOCITY AS TARGET:
βββββββββββββββββββββββββββββββββββββ
"Your velocity needs to be 40 by next quarter."
Result:
βββ Inflated estimates
βββ Cut quality
βββ Gaming metrics
βββ Trust destroyed
βββ Real velocity hidden
βββ Worse outcomes
COMPARING TEAM VELOCITIES:
βββββββββββββββββββββββββββββββββββββ
"Team A does 50 points, why can't you?"
Problem:
βββ Different estimation scales
βββ Different work types
βββ Different team sizes
βββ Points are relative, not absolute
βββ Comparison is meaningless
FOCUS INSTEAD ON:
βββ Each team's trend (improving?)
βββ Predictability (consistent?)
βββ Outcomes (value delivered?)
βββ Team health (sustainable?)
PARTIAL CREDIT:
βββββββββββββββββββββββββββββββββββββ
"Story 80% done, count 80% of points."
Problem:
βββ 80% done often means 50% work left
βββ Inflates velocity artificially
βββ Encourages partial completion
βββ Hiding incomplete work
βββ Unpredictable
Rule: Done or not done.
Not done = 0 points this sprint.
GitScrum Velocity
Velocity Features
GITSCRUM VELOCITY TOOLS
βββββββββββββββββββββββ
VELOCITY CHART:
βββββββββββββββββββββββββββββββββββββ
Reports β Velocity
Shows:
βββ Committed vs. completed per sprint
βββ Running average
βββ Trend line
βββ Variance indicator
βββ Export for reporting
SPRINT PLANNING:
βββββββββββββββββββββββββββββββββββββ
While adding to sprint:
βββ Shows running points total
βββ Compares to average velocity
βββ Warning if exceeding capacity
βββ Suggests realistic commitment
βββ Historical context
RELEASE PLANNING:
βββββββββββββββββββββββββββββββββββββ
Roadmap β Release
Based on velocity:
βββ Projected completion date
βββ Confidence range
βββ What fits in release
βββ Trade-off visibility
βββ Stakeholder planning
VELOCITY SETTINGS:
βββββββββββββββββββββββββββββββββββββ
Configure:
βββ Sprints for average (3-5 typical)
βββ Show/hide variance
βββ Include specific sprints
βββ Export data
βββ Dashboard widgets
Best Practices
For Velocity
Anti-Patterns
VELOCITY MISTAKES:
β Velocity as performance metric
β Comparing team velocities
β "Increase velocity 20%"
β Partial point counting
β Overtime to hit target
β Cutting quality for speed
β Inflating estimates
β Ignoring variance