A passionate debate about the value of estimates has erupted in the Agile community. Running under the hashtag #noestimates, advocates of ditching estimates and estimating on Agile teams argue that the estimates serve no discernible purpose other than to add overhead to the work of developing product. Even worse, they argue that estimates are overtly damaging to teams, teamwork, performance, and value delivery. Estimates become a weapon in the hands of management, which demands that teams adhere strictly to their estimates regardless of changing circumstances, new knowledge, etc.
I get it. I really do. However, I am still an advocate of estimates and estimating at the team level. Why, you might ask? (Please do!) Well, here’s the thing. Any tool – and make no mistake, estimates are simply a tool – can be misused, abused, or weaponized, but that doesn’t necessarily mean we simply stop using any and all such tools out of fear of their being abused. If that were the case, we wouldn’t have money, credit cards, the Internet, cars, electricity – you get the idea.
So back to #noestimates. I find that, with proper guidance, teams and organizations make excellent use of estimates as a planning tool. Release Planning, Sprint Planning, daily planning, and even portfolio planning all tend to work much better, be more understandable and transparent, when there are team-based estimates backing them up. It is also very clearly the case – in my experience – that teams gain a great deal from the exercise of estimating using a whole-team, consensus-based estimating technique such as affinity estimating or Planning Poker® (a registered trademark of Mountain Goat Software, LLC). And yes, I even like the traditional modified Fibonacci sequence (1,2,3,5,8,13…).
So when do I come down on the side of #noestimates? The instant the organization, management, or any other force uses the estimates as a stick to beat the team, I am done with estimates. And yes, I have recommended that teams drop the assignment of numeric estimates or any other sizing in that case. I still want the teams to engage in estimating-like discussions of work items for the learning that takes place, but the estimates can go away.
And yes, I have taken my case to management when estimates have become a stick used to beat the teams. And yes, I have been successful in helping management understand that abuse of estimates is a Very Bad Thing and have facilitated a change in behavior. I have also been invited to leave coaching engagements as a result of insisting on the appropriate use of estimates. I thought that was an appropriate instance of applying a Scrum value: Courage.
What is your experience with Agile estimating techniques and the resulting estimates? I’d love to hear from you!
All for now….