←back to thread

Reflecting on My Failure to Build a Billion-Dollar Company

(medium.com)
2101 points jamesjyu | 3 comments | 07 Feb 19 15:28 UTC | HN request time: 0s | source
Show context
holman ◴[07 Feb 19 16:22 UTC] No.19106188[source]▶
Really love this post. My favorite line is:

> Every month of less than 20% growth should have been a red flag.

I think that's pretty insightful. 20% growth is great for a normal business, of course; for a VC-backed startup it can show some warning signs about future hard decisions you might have to face.

I think there's certainly lots of discussion that has been had — and should be had — about "should I or shouldn't I raise money?", but there still are plenty of companies and founders who will raise VC, and paying attention to those early warning signs are important if that's the choice you make. It's important to worry about it each month and each week rather than the two months surrounding the raise of your next round.

replies(8): >>19106276 #>>19106282 #>>19106376 #>>19106433 #>>19106680 #>>19106916 #>>19110991 #>>19111312 #
sytelus ◴[08 Feb 19 02:35 UTC] No.19110991[source]▶
Let's run some numbers. Say you are VC putting in $1M in 100 companies. To your great luck all companies become self-sustaining lifestyle business. By definition, lifestyle business is making just enough for comfortable lifestyle of founder, so may be $300K/yr pre-tax. Let's say founder decides to give back 10% of $300K as return on investment. At that rate, it would take 33 years for VC to just recoup his original investment.

I wanted to illustrate this because it is necessary to understand why things are the way are. Lifestyle businesses is not viable for VC funding. You add on risk profile (i..e 80% of companies won't even become profitable) you get the only outcome that one or two super-hits needs to occur to cover for rest of the failures. Also remember that even with this strategy most VC funds are not even as profitable as S&P500 index.

replies(3): >>19111554 #>>19111664 #>>19112834 #
1. hn_throwaway_99 ◴[08 Feb 19 04:56 UTC] No.19111664[source]▶
Yes, but hopefully your average founder will wise up to the following: The chances of you being super successful, happy and content are much higher with a "lifestyle business" than a VC funded one.

Why?

Well, the biggest issue is that VCs can diversify (as you point out, have a lot of bets counting on a few successes), while founders can't.

I think for the vast, vast majority of people, there are greatly diminishing returns after a certain amount of money. That is, if you have a 1/10 chance of having a $10 million payout, vs a 1/1000 chance of a $100 million payout, I think most people would take the former. Put another way, most people are willing to take a good deal of risk for "fuck you" money, but fuck you money for most people is MUCH lower than what VCs expect with a unicorn.

Since VCs do always need to swing for the fences, they invariable will say no to ideas that don't have a huge potential market. My belief is then that there are a number of "mid market" businesses, i.e. ones catering to smaller niches, that have a lot of potential but have lower competition than huge, winner-take-all type businesses.

replies(1): >>19119339 #
2. chii ◴[08 Feb 19 22:35 UTC] No.19119339[source]▶
> That is, if you have a 1/10 chance of having a $10 million payout, vs a 1/1000 chance of a $100 million payout

1/10 chance of $10mil is higher expected value than 1/1000 of $100mil. Of course anybody sane will take the former and not the latter.

I think you'll find the tune different if the latter had been 1/1000 chance of $1 billion.

replies(1): >>19123737 #
3. hn_throwaway_99 ◴[09 Feb 19 18:20 UTC] No.19123737[source]▶
But again my whole point is that expected value is itself a very poor metric to use for this decision, precisely because a founder doesn't get many "swings at bat" like a VC does. Even in the case where the expected value is the same, the fact that the actual value leads to 0 for all but the very, very, very luckiest/skilled means you get St. Petersburg paradox-like results.