A friend of mine stopped by the USV office the other morning and asked me about startup mortality rates. Her business sells to startups and big companies in roughly equal measure and she told me they were seeing a rising rate of startups closing up (and thus churning as customers). While churn is never good, you really can't beat yourself up too much about your customers going out of business. She wanted to know what we are seeing.
Our portfolio has been relatively free of startup mortality in the last couple of years but it is something we expect for, plan for, and underwrite for.
I have always said this, on my blog and in person, about what we have seen at USV and what I have seen in other early stage venture funds I've worked on and invested in:
1/3 are good investments
1/3 turn into something but you wish you hadn't made the investment
1/3 are zeros
Another friend of mine asked me a similar question via email this week and pointed me to something out of Nate Silver's book, On the Edge. Nate breaks down risk in data driven ways and uses as examples poker, venture capital, crypto, etc. The book quotes some figures from Marc Andreessen wherein he outlines that a16z's funds perform as follows (note: he also broadly says these are 'top decile' fund performance numbers, too):
25% of investments make zero return (i.e. 100% write offs)
25% produce a return greater than zero but less than 1x (i.e. are losses)
25% produce a return between 1x-3x
15% produce a return between 3x-10x
10% produce a return of 10x or greater
If you bucket the first two as "zeros" or near zeros, the third one as "something you wish you hadn't invested in" and the last two as good investments, you get to roughly the same 1/3, 1/3, 1/3 that I like to use.
Of course, these are numbers over a very long period of time. Startup mortality rates rise and fall based on the vibrancy of the overall fundraising market and at times will be higher and at times will be lower.
But if you look at the data over a very long periods of time, you see that a small percentage of investments produce all of the returns. And that has always been the case for the fortyish years I have been in early stage venture capital.
Maybe that will change but it hasn't yet.

Collect this post as an NFT.
Given today's acquisition news, a reminder that entry price is hugely important in angel / venture investing. Here's a good concrete example of what a "unicorn" is worth at different entry prices. https://blog.zactownsend.com/empathy-on-entrance-price-bridge-dot-xyz-and-astranis
But what do you do with this? Pass on an investment that’s going to give you a 20x return? In most cases investors aren’t in a position to negotiate price as they’re not the lead. And even then there’s competition for the best investments which pushes up price.
You have to assume that 90% investments are a zero. So you need the outlier returns to make the IRR worth it.
That number isn’t accurate. But even assuming it’s 90% failure, does this mean you pass on the more “expensive” rounds and still get the exact same 90% failure rate?
It’s not about whether or not to pass on a 20x It’s about being weary of price points and keeping your mind sharp when looking at the terms A 20x on a seed with 100$ Million POST is statistically very unlikely while a 20X on a 2$ million seed is comparatively way more common Situations are situational and investment are situations, so the context shall be considered, which I deem to be the point Dan is eluding to
Both early and late stage companies can do well. If anything a company is more likely to succeed when already at a later stage. Which is why price is higher.
Feels like the right thing to do is to take small bets in very early stage teams. Essentially be the first check as often as you can possibly be. Even if it means you miss a lot more. Right?
Yes, aligned networking
And that’s taking into consideration that 40M post could be considered “low” for crypto startups… depending on whether you’re buying tokens or equity
These are some good examples. Did Farcaster go through a similar model for funding? Looking at such news makes it look like easy money, while 90% of such investments have little chance of being a success. Almost like investing in memecoins.
We self-funded the business for a bit, so raised at larger initial round at a higher valuation. So worse for investors.
Give me allo
Nice. Better for founders, which should also be better for the “product” as you’ll continue to nurture your baby not chase x’s primarily.
It's crazy that someone can get in with a 1K or a 7K check...
Happens all the time, esp. with RUVs now.
Ah that yes. Any deals I've even came close, there's usually a minimum (25K is most common), and even that is a hassle to manage once there are too many parties.
I've done several angel checks in that range before!
I kinda wish I did the same... Since my resources are limited, two ventures I've given checks folded quickly. Need bigger network and more bets.
Hi Casters. I had two totally separate conversations this week about startup mortality rates and so I thought I'd write up some thoughts on the topic https://avc.xyz/startup-mortality-rates
how soon does it become apparent which bucket an investment will be in?
That really depends. When I got into the business forty years ago they told me "lemons ripen quickly, pearls take longer". Which is generally true. But every situation is a bit different
How can we help startups
but their products, work for them, help them hire your amazingly talented friends, and if you have the means, invest in them
True and sometimes in the most simple form just spread the word for them