How we use Data to Ignite Action
summary
Argues that probabilistic forecasting replaces wasteful estimates by using actual cycle time and flow metrics to trigger actions: collect 10 work items from the same team, run one million simulations (Monte Carlo via ActionableAgile), pick an 85% probability range, and update continuously. A team example shows WIP limits cutting median cycle time from ~25 to 7 days, and moving a production hotfix to another team stabilizes the forecast rather than re‑estimating.
AI-written summary · about
chapters (14)
- 0:00Using data for action
- 1:07Software delivery choice
- 2:09From estimating to forecasting
- 5:21Probabilistic forecasting
- 7:07Starting probabilistic forecasting
- 10:58Forecasting with real data
- 13:45Stakeholder expectations with data
- 14:26Human side of forecasting
- 15:50Flow metrics in practice
- 16:57Cycle time insights
- 19:06Work item waste
- 20:52Aging work items and SLA
- 21:56Story points vs delivery time
- 22:38Summary
related talks
01h 01mFinding Metrics that matter and using them safely
42mData-Driven DevOps
44mMeasuring agility: Data analytics ex psychology to grow your teams agility
58mDelivering Agile BI Solutions
31mTime Theft: Using Flow Metrics to Expose Crimes from Conflicting Priorities and Unplanned Work
46mObservability for Engineering Teams
41mProficiency + Creativity = Happy Data Practitioners
28mGenerate Insights, Not Excuses: How your data is slowing you down