Every product team I have worked with prefers speed. Speed of iteration, speed of merging, speed of “we shipped it on Friday and saw the metric move on Monday.” This is mostly correct. The team that ships ten things in a quarter learns ten things; the team that ships one thing learns one. Most decisions are reversible, and the cost of being wrong is one revert.
The trap is treating every decision like it belongs in that category.
There is a smaller class of decisions that do not survive being rushed. They are usually three things at once. A name. A pricing tier. A first-time-user flow. A primary action on a chart that we will look at every day for the next year. These get encoded into the rest of the product downstream. By the time you realise the name is wrong, the SDK is shipped, the docs are translated, the customer success deck has a slide. Reverting is expensive in a way that does not show up on the sprint board.
The pattern I use now: when a decision feels load-bearing, I sleep on it. Not as a productivity habit. As a check that the half-baked version of my thinking does not get locked in before I see what the other half looks like.
The argue-both-sides test
After sleeping, can I argue both sides to the team without leaning? If yes, we usually have not done enough work on the problem yet, and the right move is more research, not a meeting. If no, the decision is ripe and we ship. The slept-on version of the choice is almost never the same as the one I would have made an hour after the conversation. Sometimes it is the opposite.
This is not a case for slowness. Most things move faster when you stop second-guessing them. It is a case for noticing the small set of decisions where one extra night of thinking is the cheapest debugging tool available.
The companies I admire have a version of this. Bezos called them one-way and two-way doors. Marty Cagan describes it as identifying the irreversible bet. The framing matters less than the discipline of slowing down for the right twenty percent of decisions, so the other eighty can run.
The work is in spotting which is which.