When one team runs several things at once
Epic/Feature Monte Carlo
A single team rarely works on just one thing. Once several Features or Epics are drawing from the same pool of throughput, "how long will ten items take?" becomes "how will throughput be split, and what does that mean for each Feature?"
Monte Carlo with Feature-level WIP built in
Feature Monte Carlo applies the same simulation principles as single-team forecasting, but adds Feature-level work in progress as a variable, so each Feature in flight gets its own range of likely completion dates and its own confidence level.
One team, several Features or Epics in parallel
Use this the moment a single team is carrying more than one Feature or Epic at a time and stakeholders want a completion forecast for each one individually, not just for the backlog as a whole.
When there's only one thing in flight, or the work spans multiple teams
If your team only has one Feature or Epic active at a time, plain single-team Monte Carlo is simpler and just as accurate. If the initiative spans more than one team, use Multi-team Monte Carlo instead.
Free Tools for Azure DevOps (ADO)
Web-based for Mac & Windows
(Power BI App)
Install the template app and connect to your data without the need to download.
InstallNot ready to connect your own data? Try the demo with dummy data →
See the maths behind Feature-level forecasting
Beyond Burnups shows exactly how Feature WIP multiplies the number of possible outcomes, and how to simulate it properly.