Michael Tromba
Michael Tromba

90% of startups fail

90% of startups fail

Then build 10 startups?

  • P(all fail) = 0.9¹⁰ = 0.35
  • P(at least one succeeds) = 1 - 0.35 = 0.65

65% chance of winning.

Jk it's actually 97%

The 65% assumes each attempt has the same 10% chance of success.

In reality, provided you are reasonably intelligent and reflective, with each startup attempt:

  • Your skills sharpen
  • Your network expands
  • You learn from mistakes
  • Your taste / opportunity radar improves
  • Your execution speed increases
  • Etc.

Maybe the win rate progression looks more like:

Startup # Success Probability Explanation
1 5% Dumb mistakes
2 8% More dumb mistakes
3 12% Even more dumb mistakes
4 18% You still suck
5 25% You suck a little less
6 32% Things starting to click
7 40% ++skills
8 48% ++skills
9 55% ++skills
10 60% Mastery

P(at least one success) = 1 - (0.95 × 0.92 × 0.88 × 0.82 × 0.75 × 0.68 × 0.6 × 0.52 × 0.45 × 0.4)

P(at least one success) = 1 - 0.0296

P(at least one success) = 97%

Should you build 10 startups?

Idk, but if you do, send me an email in 20 year and let me know how it went.