Data Volume and Sample Set Matters: Becoming a Better Venture Investor
The reality is that most venture investors and angel investors suck. They suck usually because they are arrogant or think they are smarter than you really are. It does not help that we have gone through a massive boom since 2010 and an even bigger excessive boom from 2020–2021 which has masked a lot of weakness and stupidity. So almost all early stage investors look like geniuses with the crazy good mark ups! (thank you Softbank Vision Fund & Tiger Global!). That’s changing very quickly in 2022 as the stock market and private market has imploded and the tourists have disappeared.
But I think the biggest reason that venture investors suck is because they have NOT gotten in enough repetitions to get really good.
It’s like when you meet an AI company, and they brag about how awesome their algorithm is. But then you realize how useless it is when they have NO access to data or no sample set.
The principle is relevant to being a venture investor. Like trying to do anything, you have to really suck first. Your first couple of deals are going to be terrible because you have no idea what you are doing. Add in the inflated egos coming into the VC industry, this process takes a bit longer.
For dummies like me, my 8 angel deals were complete write offs, my first 80+ deals as venture capitalist were not great. (I think I had 1 or 2 that were doing very good, the rest were meh~). My really big winners did not show up until after I learned from these. It’s like they say “There are humble investors and those that will be humbled.” It definitely helped that I did many of these investments via the accelerator program so you get pretty good feedback within 4–6 weeks working very closely with them in a 3–4 month program. This is in contrast to the much longer feedback cycles with regular seed deals. You really won’t know anything until they start growing and raising subsequent rounds and you are at arms distance most of the time. You need to do this in large numbers to finetune your radar and sense.
Now fast forward 9+ years later with over 450+ investments I’ve done and with hopefully a pretty decent track record, I can honestly say there are not that many investors out there who have my data sample set. You have to be a complete idiot (which I sometimes am) to NOT learn from this.
This sample set allows me to understand what is local maximum versus global maximum ie. what the global standard looks like. I’m able to compare all the founders I meet with what are quantitatively and qualitatively the top 10% of founders I’ve invested in and worked with. This is critical in the pre-seed and seed stage that I invest in. It allows you to compare and contrast: how does this founding team compare to the team at Shippo, Aircall, Pipefy, Manychat, RapidAPI, Embroker, Printify, Chipper Cash etc etc? It’s a great filter.
So my point is in a business as challenging and that changes as much as the venture capital business, sample set & repetition matters. You have to do MANY deals to understand what you are doing (and/or get coached & mentored by the best).
You need to love this game and be prepared to play it for a long time. Additionally, you need to approach this with a learning mindset and a large sum of humility because you will get it wrong most of the time. A great VC gets 1 big massive winner in 1 out of their 10 deal portfolio which means they are wrong 9 times out of 10.
And maybe most importantly, don’t mistake a bull market for genius! :)