The problem with a bad accelerator is that they can actually harm your company with bad advice and unnecessary distractions.
The problem with a bad accelerator is that they can actually harm your company with bad advice and unnecessary distractions.
You really should name and shame the bad accelerators. I know no one likes to be negative, but if there are accelerators out there messing up companies and leading founders astray then it would be best to make this information public. You will take some heat for doing this, but you guys should be able to handle it.
More seriously, it depends on how serious the bad accelerator problem is and you guys are one of the few that have the data on this. If it is a minor problem then I agree there are better ways to help, but if it is a major problem then it needs to be addressed.
Actually just an acknowledgement of the level of the problem with no names attached would be highly useful to the community. Very few people have access to this data and YC is one. The fact that Sam wrote about this makes me suspect that it is a major problem.
Y'know, the way it's done by all professions worthy of the name?
No, it would not be weird at all, and more importantly, it'd be good for consumers of that industry.
I mean, to make my actual argument, instead of just alluding to it, it seems to me that:
1) it's good for consumers of an industry to have ratings of the providers in that industry
2) I can think of three sources for such ratings: crowd-sourced ex-customer ratings, ratings by a purported expert-at-rating, ratings by peers. I think these each have their merits.
3) If ratings-by-peers is indeed useful, then in a given industry, it has to start somewhere.
Sometimes different alumni have very different experiences. Sometimes an accelerator's niche is very relevant to one company's problem but not another's. As Sam's advice said -- it's more useful to qualitatively talk to someone you know (and also trust, or at least consider credible) to evaluate their opinions on the accelerator. There are too many unique factors.
Obviously, this is hard for a new-ish (pre-) accelerator, so one heuristic I've used instead is to review the experience of the accelerators' founders and mentor network. It's a start.