> If you had 3 days to find bottom 50% of devs to be fired, what would be your heuristic?

Taking more than 3 days instead of rushing such an important decision.

Why does the timeline need to be so dramatically short? It doesn't need to take years or months, but executing such a change almost instantaneously with a new purchase is careless, crude, and risky.

It also reveals a lot about the mentality of the person making the decree.

Well, first of all, spend more than three days, obviously. If, for some reason, that was necessary (it clearly is not) use historical performance reviews.

> If you had 3 days to find bottom 50% of devs to be fired

I would not get myself into that situation in the first place. ;-)