I just came back from the LeanAgileUS two-day conference in Fort Lauderdale, held earlier this week. It was a great conference! One of the key themes that I noticed is that agile teams need to respond to organizational requests such as these:
- When will the next release be ready?
- How many features do you expect to be included in the next release?
Several of the presenters advocated for the use of Monte Carlo simulation, such as that found by Actionable Agile, a vendor (and conference sponsor) offering visual tools for agile teams.
Statistical PERT has always been intended to simplify the task of estimating using Excel’s built-in statistical functions. I realized that I could do a much better job of adapting Statistical PERT to the needs of agile teams who either estimate in the absence of data (using just expert opinion), or who estimate having data (such as cycle times for completed user stories).
So this begins a new quest beginning in 2018! Create useful, forecasting tools that leverage Excel and that use Statistical PERT’s simple approach to stochastic estimation.
The first result is a simple spreadsheet that lets an agile team enter a few variables based upon what they know and what they think will be true about their future efforts. Then, the spreadsheet uses Statistical PERT to build an implied, bell-shaped curve for an agile release date. Based upon the uncertainty surrounding the next release, including the work efforts of the team, the SPERT spreadsheet creates a probabilistic release date.
Try it out! Go to the Download page on this Beta Blog website to download development-only versions of Statistical PERT for use with agile teams.