Title: No Estimates: Alternatives to Estimates
Convener: Woody Zuill
Participants: There were approx. 45 participants
Discussion: We asked the question: Why do we think we need estimates?, and collected post-it notes with the reasons we “think we need estimates”
We then grouped them, and found the following groupings:
1 – Predictability (about 3/4 of the notes collected)
2 – “Go or No Go” decisions
3 – Prioritization
4 – “Fill” a Sprint with work
This led to a second rhetorical question: Why do we need predictability?
- We decided that a “5 whys” exercise should be done to get down to the bottom of the real reasons we insist that estimates are needed. We discussed the idea that MUCH of the training in the Agile field is about estimates, prioritization, sprint planning, release planning and so on, and yet there is nothing in the Agile Manifesto that points to the idea that estimates or this type of planning is needed or important. At this point we asked for discussion on possible alternatives to estimates.
A few were discussed:
- Pull discrete pieces of work
- Early and often delivery of working software
Some time was spent addressing the idea that “we still have to do estimates to be able to make software”.
Conlusion: While there are some people working without estimates, there are still many unanswered questions about how to go about doing that. It was observed that there are many different environments and while “no estimates” might work in some, it might not work in others. In other words, we had no “conclusion”, but we did have a lot of ideas we can explore when we get back with our teams and daily work.