So I’ve been investing long enough — just about 10 years — to not really have mastered it, but seen the results of trade-offs made in investing. I’ve made about 30 material investments. More on them here. Most have been successful, but of course, not all have, and I’ve been reflecting on the ones that didn’t, because they still had attractive elements when I invested. I’ve at least come far enough to see why some exited for billions (Salesloft, Pipedrive, Greenhouse) or become truly worth billions with hundreds of millions of revenue (Talkdesk, Algolia, etc.) … and others didn’t. Despite strong initial promise.
What Works — That You Might Think Wouldn’t Work:
- High employee/VP churn. You’d think turning over lots of VPs would impact growth — and it does. But some CEOs get through this as long as they are super committed to bringing in the next group of VPs. This can be especially common during hyper growth.
- Solo founder, too many founders, co-CEOs. Solo founders can work (Zoom proves this). 5+ cofounders can work (I think of that as too many cooks, but now I see it still can work). Co-CEOs can work, even though many investors think this is a flag and confusing. It works for Atlassian. It works for others, too.
- Taking a long time to get to $1m-$2m ARR. I used to think if you didn’t get to Initial Traction fast enough, the team would burn out. But now I’ve learned that’s not always the case. Some of my best investments had zero revenue the first 2 years or even longer. UiPath took 10 years to get to $1m in ARR!
- Cofounder conflict. I hate to see it, and personally, it held me back. But now I’ve seen many unicorns make it even with significant cofounder conflict in the early and middle days. Some of the best CEOs just push through it, one way or another.
Signs, With Hindsight, Of An Investment That Probably Won’t Work Out:
- CEO hid things and/or was misleading. If the metrics don’t make sense, just don’t invest. I can think of one exception that is a unicorn now, but otherwise, if the metrics are a bit baloney (e.g., claiming bookings are ARR, or using Quarterly MRR, or claiming team members are full-time that aren’t) … then pass.
- If a CEO surprises you with things, do not invest. Hiding the ball, I’ve seen 100% of the time, leads to a mediocre outcome. Not always a failure, but always a mediocre outcome. This is really just the prior point amplified. The best CEOs are direct with the good, the bad, and the ugly. At least by the second meeting.
- Great CEO But Mediocre CTO. Sometimes, you can grow quickly at $1m-$2m+ ARR, even with a mediocre CTO. Because that one 10x feature might be enough up to that point. But then … things get complicated. You have to scale, and add 10x the workflows. A mediocre CTO can’t keep up. These ones, even with a great CEO, hit a wall somewhere. It may be as late as $10m-$15m ARR, but somewhere. I’ve seen some of these still scaling to real value (hundreds of millions), but I haven’t seen a mediocre CTO take a startup to the stratosphere yet in SaaS. Even if the CEO is great.
Things That Are Super Risky You Might Not Think Are: These are flags of likely issues to come, but they aren’t dispositive.
- Taking the First VC Money Offered Too Quickly, Especially if It’s a Suboptimal VC. The best CEOs take their time. The ones that immediately take the first term sheet offered out of a bit of panic, I’ve seen that decreases the odds of success. It doesn’t kill a startup, but it does turn out for me at least to be a sign of a CEO that is too worried, nervous or conservative. Take an extra week or two to make sure the investor you think you want is the right one.
- Secondary Liquidity Too Early. Selling some founder’s shares later, as the valuation passes $100m, makes a ton of sense. It helps you go long as a founder. I should have done it. But selling too early, at too low a valuation, is a risky sign the founders don’t 100% believe. This isn’t 100% correlated to failure, but there is a strong correlation here in my experience. It also can lead to substantial co-founder conflict down the road when an underperforming co-founder expects more and more of their stock to be cashed out. A lot of experiments were run here in the Go-Go Days of 2021. They didn’t work out.
- A Burn Rate Even a Smidge Higher Than Normal. This is super risky — because it only grows from there. A burn rate that is even 30%-40% higher than similar companies is a flag. It’s a flag the burn rate will continue to expand at this rate. For me, any start-up that has burned more than say, $3m or so on the way to $1m+ in ARR–that’s too much. It’s a sign they just need to burn too much for each new $1 they bring in. The meta-learning is more money makes it worse. They ratchet up the burn, and then burn even more than similarly situated companies. Even a slightly-higher-than-normal burn rate compounds. Into a too-high burn rate. Almost every single time in my experience over the past 10 years. More here.