Gut Feel Estimation
Joshua Kerievsky ????
Helping organizations deliver better software sooner | Dad | CEO | Entrepreneur | Author | International Speaker | Software Designer | Tennis Player
In 2008, a small Silicon Valley company (that would later be acquired by IBM) began their agile journey with a single pilot team, trained and coached by Industrial Logic. The pilot team engagement had gone well and now the executives wanted more teams to learn to be agile. I wanted to interview the team leader, a talented product manager, before we devised an agile scaling strategy for this company.
I met the pilot team's product manager in her team's room. There were agile posters everywhere, including iteration plans, burn down charts, release plans, retrospective boards.
I asked her how she liked the agile process. She said she liked it. I asked how her team liked it. She said they like it. I asked many questions but I felt like she was holding back. As I looked at the posters on the wall, something seemed too perfect and it led me to ask about estimation.
Me: So how's the estimation process working for you? Do you and the team like story points and tracking velocity? (My own team had given up using sprints and story points a year earlier).
Her: Yeah, it's working fine for us.
Me: So you're able to estimate accurately with story points and get a consistent velocity?
Her: Yeah.
Me (feeling she was really holding back): Are you sure?
Her (hesitating and looking me over closely): Um, you want the truth?
Me: Yes I want the truth!
Her: You can't handle the truth! (No, she didn't say that, but I'm sure you were thinking it!)
Her: Well, we don't actually use story points or track velocity.
Me (surprised and finally feeling like she was opening up): But what about those posters on the wall? You seem to be tracking that stuff on the posters, no?
Her: Oh those, well the truth is we made those posters because you were coming to visit us today!
Me (stunned): Really?!?
Her: Yes, to be honest, we wanted you to give management a good report about how agile were are.
Me: Really?! So what do you actually do for estimation instead?
Her: Well, we get together at the beginning of an iteration, the team does gut feel estimates of what we think we can accomplish during the iteration, and then we get busy doing our work.
Me: And what happens when you don't quite get all of the work done in an iteration?
Her: It's no big deal. We just do more gut feel estimates for the next iteration. We're usually pretty accurate in our estimates, so it's never been a big problem.
Me: There's nothing wrong with that approach whatsoever! It's perfectly agile to not use story points or velocity! In fact, we don't use those things in our shop and we dropped iterations (sprints) as well!
Her (looking quite relieved): Oh, that's wonderful to hear. I thought you were coming here to check up on us and see if we were practicing agile exactly as your coach taught us. I've been doing this kind of work for a long time and some of the agile stuff you taught us didn't feel natural to our team, so we didn't do it.
Me: That's perfectly, perfectly fine! You've gotten excellent results and produced a working product quickly. Thank you for your time!
I went back to the executives and raved about how great this product manager was, and what an excellent job her team had done.
As we helped more teams learn to be agile across this company, we didn't teach them story points and velocity calculations. We encouraged them to estimate in their preferred way and to focus on managing risks and evolving their products over successive iterations to become releasable. (Several years later we'd drop even teaching iterations). This approach worked well: we didn't get the normal confusion around story points, or teams beating themselves up for not having a consistent velocity or managers comparing teams by their velocity or other silly stuff.
Gut feel estimation is primarily what my colleagues and I use when we develop our own software. We always aim to work small, so if our estimates exceed 3-4 days for any piece of work, we'll find a way to de-scope that work. The key is that all voices are heard with respect to estimates.
Dan North wrote about his experiences with gut feel estimation in a blog called Blink Estimation.
Have you experimented with different ways to estimate on your agile projects? What feels most natural to you?
Senior Engineering Manager at HackerOne
7 年In essence I normally don't really care what the teams use as estimation ...as long as it works for them. Even if it would be how many beers does it take to develop something ;). I even heard about teams that just gut estimated the number of stories they could pick up..and they were almost always spot on because the refined it to a level they had a good understanding about what needed to be done. I do think there has to be some kind of estimation... regardless of the method used..
Agile Coach
7 年Great article. Teams get soo caught up in points and accuracy. I do a hybrid approach where I do reality checks during and at the end of planning without looking at the accumulated story points. It reminds them to always take a step back and look at the bigger picture. I also do gut feeling and reality checks towards the 2nd half of the sprint.
Empowering Organizations through Systemic Value Creation, Clear Communication, and Agile Principles"
7 年I perfectly understand the product manager, as dealing with estimates and velocities can be a source of unhappiness, specially when too much conclusions are made from it. However I do find value using it as a way to drive a planning session discussion, even if you only use T-shirt sizing.
Cofounding Partner and CEO - TRG Technology Revolution Group
7 年Thanks Joshua for your post, in our agile project we were taught burn down charts during the induction but we never used them, one burn down chart was hanging on a wall but dust soon built up on it... What we did was a grooming meeting every week where the purpose was not as much to estimate story points (we used planning poker and Fibonacci numbers) but to really understand what was the work involved under the user story.
Creating Learning Organizations through Change & Leadership
7 年Thanks for sharing. Whatever technique you use as an estimation, I have the feeling that teams and organisation always get too hung up on the numbers instead of the value they would like to create. Therefore I started to coach teams to use story point estimations not to know, how long / how complex / how big an activity is, but rather to ensure, that all included members do have the same understanding of what they want to achieve. Then, the velocity resembles the result of the sprint (or a series of sprints) and not the outcome of a planning.