Forecasting & Release Planning
Forecasting and release planning test how Scrum handles uncertainty. The exam expects forecasts based on evidence, not promises, and release plans that can adapt.
Core rules
- A forecast is a probabilistic statement, not a guarantee.
- Use past data such as throughput, cycle time, and Sprint results to forecast.
- Keep the Product Goal and value as the guide; release scope may change.
- Make uncertainty visible; provide ranges rather than single dates.
- Reforecast as evidence changes; do not protect an old plan at the expense of value.
Worked example
A stakeholder demands a fixed release date for 40 backlog items. The team has historical throughput of 6 to 10 items per Sprint. The Scrum Master helps the Product Owner show a range: likely completion in 4 to 7 Sprints, with scope options for the fixed date. Reasoning: the forecast uses evidence and exposes trade-offs instead of turning velocity into a commitment.
Traps
Treating velocity as a productivity score or promise; committing to scope, date, and quality simultaneously; forecasting from wishful capacity; ignoring dependencies; using story points across teams as a comparison.