Survivorship bias

[byrongibson]

Hi Dan, I just came across your book page from a Hacker News post (http://news.ycombinator.com/item?id=2857184), looks like an interesting project. Just a thought, in case this hasn't hit your radar yet, don't forget to take into account survivorship bias.

After gathering the list of patterns common among successful startups, you also should attempt to filter them by the degree to which they were present in failed startups as well. Not an easy task, tracking down info on organizations that no longer exist. However, one possible strategy might be to show the final list to experienced VCs, angels, and serial entrepreneurs, and ask if they noticed any of those patterns in their failed investments.

Additionally, it might be instructive attempting the inverse as well - investigate the patterns common to failed startups (and possibly to what extent they exist in successful ones, eg do strong positive traits compensate for negative ones, etc.).

Fwiw, one of the biggest failures of social science researchers to account for survivorship bias was the book Built To Last, in which researchers tried to identify the traits that companies that have existed for decades all share. But they forgot to take account for the fact that companies that failed also possessed some of those traits:

http://en.wikipedia.org/wiki/Built_to_Last%3A_Successful_Habits_of_Visionary_Companies

Good luck, will keep an eye on it!

Byron

[dshipper]

Hey Byron,

Thanks so much for reaching out I appreciate it!

You make a very good point. I had never really thought about it, but I guess
you're right that many successful companies may have elements in them that
have impeded their growth even though they may have overcome them. I think
it would be really cool to go into the things that failed startups share,
but I have a feeling that it's something like:

• They didn't solve a problem that anyone wanted to pay money to solve
• They couldn't cost-effectively reach their customers through existing
  sales channels
• They raised too much money, and expanded to fast, missed their wildly
  optimistic revenue projections and imploded

Still an interesting thing to look at since in my opinion a startup is never
really dead until the founder decides it's dead. Thanks for the book
recommendation too :) I am 100% going to read that.

Best,
Dan

On Sun, Aug 7, 2011 at 8:22 PM, byrongibson <
reply@reply.github.com>wrote:

Hi Dan, I just came across your book page from a Hacker News post (
http://news.ycombinator.com/item?id=2857184), looks like an interesting
project. Just a thought, in case this hasn't hit your radar yet, don't
forget to take into account survivorship bias.

After gathering the list of patterns common among successful startups, you
also should attempt to them by the degree to which they were present in
failed startups as well. Not an easy task, tracking down info on
organizations that no longer exist. However, one possible strategy might be
to show the final list to experienced VCs, angels, and serial entrepreneurs,
and ask if they noticed any of those patterns in their failed investments.

Additionally, it might be instructive attempting the inverse as well -
investigate the patterns common to failed startups (and possibly to what
extent they exist in successful ones, eg do strong positive traits
compensate for negative ones, etc.).

Fwiw, one of the biggest failures of social science researchers to account
for survivorship bias was the book Built To Last, in which researchers tried
to identify the traits that companies that have existed for decades all
share. But they forgot to take account for the fact that companies that
failed also possessed some of those traits:

http://en.wikipedia.org/wiki/Built_to_Last%3A_Successful_Habits_of_Visionary_Companies

Good luck, will keep an eye on it!

Byron

Reply to this email directly or view it on GitHub:
#1

-Dan Shipper