True super-early stage investing done right should have a high failure rate.
As an investor, including an accelerator, it’s actually surprisingly early to have a low failure rate. Just invest in every start-up with some traction, a low burn rate, and reasonably happy customers. Those ones don’t fail that often. And when they do fail, it’s usually very slowly.
I’ve done OK as an investor so far, but even I’m a bit guilty here. Of 23 investments, 0 have failed so far. But then again, only 2 of the 23 were pre-revenue and pre-product market fit. And those 2 were founders I personally knew well.
The problem is that really safe investments don’t always, or maybe even almost ever, generate the insane venture returns. Dropbox, Airbnb, Coinbase, etc. all were quirky and pre-revenue when they entered YC.
A high failure rate is part of a high return rate in pre-seed investing.