Many years ago I read an article about a venture capitalist (I think) who took over the ailing Avis car rental company. He said that the most important thing he did to turn around the company was to fix a deadline of one month for all decisions. If you have been trying to work out which of several options to choose and have spent more than a month on it then you just toss a coin.

The rationale is that no amount of analysis will guarantee the right choice so more time is just a waste when you could be finding out in real life whether or not your decision was right. Instead of spending six months agonizing over the decision and still making the wrong one you get to make that wrong decision five months earlier and discover a month later that it was wrong. Now you are four months ahead and can go for one of the other options.

Makes sense. In general, I found there is a middle ground. I often look back and I am thankful to myself for "marinating" on something a bit, as "jumping in" has its drawbacks.

You can spend weeks implementing something and then having to start from scratch - time wasted. It's important to at least play certain decisions forward in your head, understanding that you will not know a lot in advance until you actually run into them - discovered work.

Sometimes I wake up in the morning and a bulb goes off - "oh, THAT would be a huge problem! Next option."

So yes, marinate, but up to a point.