I looked at stats available on: techstars.com/companies/stats/ and saw they have a success rate (Active + Acquired companies) of over 90%, but I only see 75%. Here's why.

The first thing I did is calculate without their last 10 cohorts. The reason is these cohorts graduated less than 1 year ago and haven't had time to succeed or fail yet. Without the last 10 cohorts (111 of their 291 startups) the success rate falls to 85% when they are removed. (Or down to 80% is you are more realistic and remove any startups younger than 2 years, but I went conservative.) I'm not changing their definition of success as being survival rate, rather than some other metric.

The second thing I did was look at all startups marked Active. WIthout a way to judge if they are Active other than what I see online, as a proxy I looked at whether they had posted to a blog, Twitter, or Facebook anytime in the last year. In some cases there was news of them going into the deadpool or a domain for sale, but for the most part they look like Zombies, not officially dead.

It's not a perfect way to judge so please let me know if any of these startups that look inactive are really alive: J-Squared Media, Application Experts, UsingMiles, HaveMyShift, InvitedHome, ScriptPad, Marginize (the team formed a different company but I'm counting this one as failed), Highlighter, RewardsForce, Wantworthy, Creative Brain Studios, Strohl Medical Technologies, Veri, CloudSnap, Emergent One, Rewind Me, ReplySend, RollSale.

To measure accelerator programs against this 90% standard, I'd want it to be accurate. Again, I like TechStars and I'm posting this to learn. If we measure all accelerators the same way, I'm sure a lot of obscure programs have high success rates as well.

Thanks.