Bet On Slope - Chris Yin

A couple years ago when I was running product at RainforestQA, the phrase "hire for slope and not y-intercept" became popular. I'm not sure where it came from (Sam Altman seems to think it was Gmail creator Paul Buchheit), but it was pithy and quickly became common startup parlance.

The point was that if you are hiring and evaluating 2 difference people for a role -- one with more experience and the other with less, at first blush it seems obvious that you should hire the one with more experience.

However, that decision is too fixated on a single point in time. If you zoom out, the point is that the hiring criteria should be less about experience, and more about growth rate. Someone with less experience but growing quickly will eventually outpace someone with more experience but stuck in their ways.

Put another way -- in this shitty diagram I drew, you can see that you should probably hire person a over person b.

What I've realized is that this model doesn't just apply to hiring, but it's a simple framework that fits most decisions if you are looking for outsized returns.

As a founder, your primary job is to pick the right market. And the way to pick the right market is by filtering one that is not huge yet today, but growing very quickly. Read Chris Dixon's post what the smartest people do on the weekend is what everyone else will do during the week in ten years for more. The crux is the same -- a smaller but growing market is better than a large static existing market (for founders).

As an early stage investor, your job is to pick not only the right markets, but also the right companies (which imo means picking the right founders). The startup graveyard is littered with companies where the "right founder" -- tier 1 pedigree, tons of experience doing that exact thing, etc -- gets smoked by a raw founder that knew nothing but was super high throughput. Another way I like is how Ben Horowitz phrases it -- look for strength and not lack of weakness.

Even on the other side of hiring, as an employee you also want to filter by companies that are high growth. The glory that you get from early stage startup building is only a thing before its obvious. Once it's a known thing, you are too late. And so you want to filter for companies that are both in the right markets but also growing quickly. Usually once a measurable growth rate is available it's already too late. And so the way to judge growth before that is by speed of execution. Join those that follow the 100 pots philosophy.

At Plume we've lived by and been the beneficiary of this idea across every area. When we started, nobody cared about RWAs. It was seen as a dead market and in fact most people tried to convince us out of it (we even seriously considered doing a BTC L2 instead, which would have been a massive mistake). Not even 2 years later, it is the dominant narrative across crypto and now everyone is trying to work with us.

Our early team was the same. Many were very young or without major accomplishments. Low y-intercept. Fast forward to today many of them are highly sought after and at the top of their fields. And investors that wouldn't even meet with us are now desperately tweeting about RWAs and making bets into low quality projects to gain some relevancy. They did not bet on slope.

Now there are obviously many cases where it's not all about filtering purely by growth rate. Generally for later stages where you are focused more on stability, predictability, maintenance, etc then hiring for immediate experience might be better. Classic examples are finance or compliance (although I'd argue it's less about focusing on y-intercept but that the required slope to be competitive here is just really high and thus hard to hit).

Another example is in the very late stage or public markets. In the public markets there are so many participants and potential distortions that can throw things off that purely betting on slope can sometimes still lead you down the wrong path. But similarly, I would argue that it's still roughly right but mostly about time frame -- over the long arc, growth rate still wins out even in public markets but you do have to survive long enough for that to show itself.

And then lastly the most obvious comment is that when you bet on slope over y-intercept you will sometimes be wrong. Maybe you over estimated the slope or something changed and now it's slowing down. By nature these decisions are riskier than the alternative. But that's why I phrased it as a framework for those looking for outsized returns -- you don't get those by doing what everyone else does. If you are not focused on outsized returns, y-intercept might just be fine.