Hacker News new | ask | show | jobs
by eli_gottlieb 5191 days ago
9/10 start-ups fail, just like most other forms of new business. I believe we could apply Bayes' Law here...
2 comments

But that doesn't mean 9/10 of startup founders fail. Most of the successful startups I know today are not the founder's first startup. Everybody says doing a startup, even (especially?) if it fails, is extremely educational.

The most common pattern I see is doing a startup, failing, but learning enough about how to do it right, and then trying again and having moderate success with #2 or #3.

9/10 of startups failing could be explained by "virtually all first-time startups fail, and 9/10 founders give up after their first failure, and those that stick with it, succeed". That would actually be rather encouraging.

Almost nobody I know working for a big company is working for their first big company, so you could also say "9/10 of all employments fail", but that's hardly an indictment of employment.

Yes, but much as 9/10 employees won't ever make 7 figures, 9/10 startup founders won't ever cash out for 7 figures. Plus, in a cycle of invest-boom-invest-bust, your previous successes can be wiped out, making stable growth more difficult.
That is crazy talk. The rate for small businesses as a whole is around 50% failure within the first 5 years.

There are exceptions to that rule and ways to increase the odds. As an example, 90% of franchise businesses are still running after 5 years.

There are a lot of stupid numbers flying around. The problem is defining "business" & "failure."

Some businesses are only meant to last a couple of years. Lots of owners start several businesses and eventually focus on one. Most numbers you hear are based on the number of businesses registered & renewed/not renewed, a terrible way of measuring both these things.

Jaxn's data isn't anecdotal -- those numbers were reported in the WSJ earlier this year in an article on franchising.
That's true. They are also in my deck ;)
Anecdata: Walk down a main commercial street in any large metropolis and count the number of completely unviable businesses (these are the ones that turn over their real estate every few months). These are further polluting the statistics.
I am not knowledgeable enough to know whether or not a business is viable just from looking in the storefront.

Though I would argue that as the owner of two retail locations, I am probably better equipped to make that judgement than most.

In my experience, most retail leases are for 5 year terms. I can't imagine that the default rate on those is anywhere near 50%. Most business "failures" likely never got far enough to sign a lease.

My gist was you don't have to be able to know, just walk down the street, then walk down the same street 6 months later, and many of the shops will have changed.
If you're just looking at storefronts, how do you know what's viable and what isn't without making assumptions? Looking at real estate turnover tells you something about the location - and what businesses are appropriate for it - but doesn't tell you much about the viability of any particular business model.