Scrum Was Built For Scarce Engineering Time

Scrum was built for a world where engineering time is scarce, and I don't think we live in that world anymore.

If you only have so many developer hours, it makes sense to plan ahead. You refine, you run sprint planning, and you try to make sure the next two weeks go into the right things. With AI tooling we have a lot more capacity than before, in some areas more than we can use, so the planning protects a thing that is no longer short.

Before, a good idea waited one or two weeks for sprint planning, and once it was picked it launched about a week later. Often five weeks passed before anybody knew if it worked.

Today you give the idea to Claude, review what comes back, and it is deployed the same day or the day after. Two or three days later the dashboard shows whether the idea was any good. So instead of three meetings with five stakeholders who all have an opinion, we discuss actual data.

We stopped Scrum almost a year ago and moved to a Kanban style. Every problem gets a small team of a product manager, a designer and two or three engineers, and the number I care about most is the time from an idea to the first real numbers.

Big bets are different. A new integration or a large refactor across the product still gets planned and aligned up front.