The Last Responsible Moment

The best time to discuss — how are we going to do this? — is just before doing this.

Agile has borrowed some great ideas from Lean. Limiting WIP (work in progress) is the most notorious. Just in Time is another, and it is this concept that I want to cover here. I have heard agilists use the term last responsible moment, and it is the same idea expressed in different verbiage (which I prefer).

This blog is not to debate estimation vs. #NoEstimates. It has advice for teams that are still using old school estimation and aims to put to bed the argument for dismissing #NoEstimates or other forecasting mechanisms on the ground of not having the valuable “how are we going to do this?” conversation that a story pointing/velocity team will have when asked to size a work item.

Waste

Having the “how are we going to do this?” conversation too far in advance of starting the work comes with a high chance of that conversation turning into waste (another lean concept). Here are three examples of why/how this conversation becomes waste:
1 — the work may never happen because conditions have changed and it no longer makes no sense to do it
2 — the work may be out so far out in the future that the assumptions made are no longer true once the work does get scheduled, e.g., the architecture and/or technology has changed, resulting in having to have this conversation all over again
3 — the conversation happened long enough ago that the team forgot it or key aspects of it, and they need to have the discussion all over again
4 — the team makeup changed since the conversation happened, and the views are now different

Advice for Teams Still Doing Old School Estimation (Scrum + Story Points + Velocity + Refinement)

If you have been asked to size a work item and the conversation descends into an architectural discussion, let it run for, say, 30 seconds or a minute (pick a short timebox). If not resolved, stop the conversation. Ask for everyone’s estimate, average them and call it done. Only have the in-depth conversation just before the work is going to get scheduled. If the estimate changes, so be it — adjust the work scheduled for the upcoming iteration.

Software Developers Time and Energy are Taxed by Estimation and Planning — Don’t Abuse It

Estimation comes with a high tax on developers. Don’t believe me? Go and ask a software developer. It taxes their time and creative energy, both of which are very expensive/valuable. Don’t go throwing that away, i.e., waste it. (There is the waste word again.)

Summary

There is a lot of value in a team discussing how they will tackle a work item, but that value diminishes, and cost increases when the conversation happens too far in advance of the work being scheduled. (A common anti-pattern in reductionist/determinist models of work.)

In alternative forecasting methods to the now-traditional story pointing and velocity method, the “how should we do this” conversation still happens and is not lost. Please do not dismiss alternative forecasting ideas for fear of losing this conversation. Keep your mind and options open. Keep exploring and experimenting.

The agile manifesto states that “We are uncovering better ways of developing software.” Uncovered is static. Uncovering is not static and is ongoing. Let’s stay agile and keep exploring.