SAFEs vs. priced rounds, revisited.

The case for a SAFE has always been speed and cost. The case against it is that neither of those advantages is real once a company has stacked three or four of them at different caps and is trying to work out who owns what. The question is not which instrument is better. It is at what point the deferred complexity becomes more expensive than the friction it was meant to avoid.

For a first cheque into a company with no revenue and no governance to speak of, a capped SAFE remains the right answer. There is nothing to price, and a priced round would impose a board, a set of protective provisions and a diligence exercise that the company cannot yet justify. Below roughly a quarter of the eventual round size, the maths favours speed.

The crossover comes when the aggregate of outstanding instruments starts to determine control rather than merely dilution. Once converting SAFEs will hand a meaningful minority to investors who have never negotiated a governance term with the founders, the next lead will insist on cleaning it up — and the cleanup happens at the founders' expense, because it is negotiated from inside a live round rather than ahead of one.

The practical test is whether the founders can state, without opening a spreadsheet, what the cap table looks like on conversion at three different valuations. If they cannot, the round should be priced. The friction is the point: it forces the conversation that the SAFE stack was deferring.