#noestimates doesn’t work…
I was interviewing a potential engineer today and asked them about their philosophy about estimates. It seems odd that a person can have a philosophy about something as simple as estimating work or effort, but it’s a totally valid question in the tech space.
Their answer suited my needs… they are OK with estimating effort but tended to buffer when providing a timeframe. They also mentioned that when working with other engineers they added more buffer because no one seems to get it right. The second part I found funny because somehow they thought their estimating ability was superior to others.
At least they understood two fundamental things:
Estimates are needed when prioritizing work so that value/effort ROI can be calculated and understood
Estimates are often wrong because humans aren’t great at it
How to get better estimates
I’m not the first or last product person who wants estimates and I’m not here to explain why they are needed. I want to share how you can overcome #2 above, the inevitable reality that estimates are going to be wrong because humans are fallible.
Explain why estimates are important to your company. Simon Sinek has taught us to always start with why. Invest some time in ensuring that the engineers understand the value of knowing when a feature will be available not only to the sales team, but the customers. If you can tie estimates to customer empathy they will be less resistant.
Talk with the developers about how they like to estimate. Do they prefer to estimate using points? weeks? sprints? hours? t-shirt sizes? something esoteric? By allowing them to be part of the decision, they will start out with more buy in.
Establish some rules and limits. Agree to things like if an estimate is larger than 2 weeks that the request needs to be broken into smaller pieces. Or that if engineering estimates across the team are too varied that some time-boxed dev research needs to be done. Empower the entire team to create and enforce the rules.
Look back to look forward. Retrospectives aren’t only about what went well with the team, at strategic points in a project look back and see how on target the estimates were and discuss options for improving accuracy in the future. When you do this, don’t forget to remind people why estimates are important in the first place.
What have you done to improve the estimation process at your company?