The Investments Where I’m Going to Lose All My Money

by Jason Lemkin | Blog Posts, Exit Strategy, Fundraising

So when I started writing venture checks in 2013, I didn’t know what I was doing, but I had a strong start:

First was Pipedrive co-leading seed, then acquired for $1.5B cash

Second was Algolia leading U.S. seed, now at $200m+ ARR and an IPO candidate

Third was Greenouse/Parklet, acquired for $800m and now at $200m ARR

Fourth was Salesloft, acquired for $2.5 Billion cash

Fifth was Logikcull, acquired for $300m cash

So it was a good start. Since then, I’ve made some pretty good other investments as well. But also, I made some that … weren’t.

For a while, the 2021 Go Go Days masked everything. Almost everything looked great and had an up-round. 2024 unmasked a lot of things. And now, the AI Boom is masking things once again,.

Now as an investor in startups you are going to lose money on deals. It’s OK at some level. It’s part of the model. But now I can see clearly why in some deals, I’m going to lose all my money.

Reasons I've lost money investing, or had mediocre outcome:

– founder wars
– CEO misled on metrics
– CEO ignored zero cash date

What didn't stop unicorn outcomes:

– one terrible year
– bigger competitors
– taking a long time

— Jason ✨👾SaaStr.Ai✨ Lemkin (@jasonlk) July 9, 2025

The top reasons an investment has turned out to be a Zero:

#1. Any misrepresentation about the financials, no matter how small

If the financials are misrepresented even a smidge, I’ve lost all my money or almost all. If revenue was overstated a bit. If contracts were claimed to be closed that weren’t quite closed. If “ARR” turned out to include a bunch of one-time services revenue. Does it really matter? I think it turns out that it does. It just gets worse. The small misrepresentation at the seed becomes a bigger one at the A, becomes a fatal one at the B. I’m going to lose all my money here. Every single time.

#2. If founders hid anything in the round and/or in diligence. Almost no matter how trivial

Related to but not quite the same as #1. If the founders hide churn, or hide a co-founder is leaving, or hide that a key customer just gave notice, or really anything that much matters, again, it’s just a smoking gun. It gets worse. They’ll hide even more. Founders who hide things at the seed hide bigger things at the A, and by the C round they’ve stopped telling the board the truth at all. The pattern always plays out the same way.

#3. If the founders refuse to get the burn rate under control

OK these ones aren’t always zeros. But they’ve been close to it in my experience. I haven’t done better than 3x here. Note this isn’t just a unicorn issue. It can happen after raising just a few million, too. The 2021 era trained a generation of founders that burn was a feature, not a bug. 2024 retrained most of them. The ones who didn’t get retrained are the ones I’m losing money on. In 2026, with AI letting lean teams do 5-10x what bloated teams did in 2021, there’s just no excuse for refusing to right-size.

#4. If I relied on anyone else being in the round or having already invested, unless they said it was clearly the best deal they are in right now

I haven’t done an investment just because a great investor was already in the deal, but I’ve almost done it. I’ve skipped some steps because of it. Here, I will lose all my money. You have to ignore who else is in the deal. Tier-1 funds are wrong all the time. They’re allocating off a portfolio model, not betting their net worth. You’re betting yours. The only signal that actually matters is when a great investor says “this is the best deal in my portfolio right now.” Anything short of that is just noise.

#5. Not Actually B2B. Pseudo-B2B but Not Actually B2B

I won’t do any of these anymore. The B2B playbook just doesn’t really work in pseudo-B2B. Or rather, it seems to work at first, but then it doesn’t really scale. The gross margins look fine until they don’t. The retention looks fine until you realize half the “subscribers” are really one-time buyers on a payment plan. The CAC payback works until the channel saturates. And in the AI era, this is even more important to get right, because a lot of “AI B2B” companies are actually AI services companies with a recurring invoice. Not the same thing. The unit economics will betray you eventually.

Now this sounds dramatic, and it is a bit dramatic, but the winners will more than balance out the losses. In aggregate, these total or almost total losses won’t add up to a huge amount when measured against the gains and winners. A real amount, but not a huge amount. That’s the math of venture, and the math has held up.

But these losses were 100% avoidable. That’s what sticks with you. Every single one of them had a tell at the time. Every single one. I just didn’t listen hard enough.

The lesson, 13+ years and 2+ funds in: the red flags are almost always visible at the seed. Trust them.

The Investments Where I’m Going to Lose All My Money