The Startup Massacre: 966 Companies Dead

While everyone was busy arguing about whether AGI would arrive in 2026 or 2027, something way more interesting happened: 966 venture-backed startups quietly shut down in 2024.

That's not a typo. Nine hundred and sixty-six.

According to Carta's data, that's a 25.6% increase from 2023's death toll of 769 companies. AngelList reported an even more dramatic 56.2% jump in wind downs year-over-year. And here's the kicker—we're probably undercounting. 

These numbers only capture companies that formally dissolved through platforms like Carta. Plenty more just... stopped returning calls and faded into the ether.

I spent the last few weeks digging through every major shutdown of 2024. Builder.ai burning through $445M on fake AI. Canoo's $1B+ EV implosion. Lilium's eVTOL dreams crashing back to earth. And dozens of others you've never heard of.

What I found wasn't random bad luck. It was four specific patterns that killed nearly 1,000 companies. And if you're running a startup in 2026, you need to know if you're following the same playbook that just murdered a thousand of your peers.

Over the next four weeks, I'm breaking down each pattern. Today? The one that looks like success right up until everything collapses.

The Numbers Don't Lie (But They Tell an Uncomfortable Story)

Let's establish what we're dealing with.

The 2024 carnage by the numbers:

• 966 U.S. startups shut down (Carta data)

• Shutdowns increased at every stage: Seed up 102%, Series A up 61%, Series B up 133%

• 74% of shutdowns were pre-seed or seed stage companies

• Finance led the casualties at 15% of total shutdowns, followed by food (12%) and healthcare (11%)

But here's what nobody's talking about: This isn't a 2024 problem. It's a 2021 problem finally coming home to roost.

During the zero-interest-rate bonanza of 2020-2021, VCs threw money at anything with a pulse and a Figma mockup. 

Companies that had no business raising $50M seed rounds raised them anyway. Valuations got completely detached from reality. 

Due diligence became a joke—some deals closed in 72 hours with barely a phone call.

Peter Walker from Carta put it perfectly: "The working hypothesis is that VCs as an asset class did not get better at picking winners in 2021. In fact, the hit rate may end up being worse that year since everything was so frenzied."

Translation: We funded a shitload of companies that should never have gotten funded, and now they're all dying at the same time.

The time lag is about 3 years. Most startups raised big rounds in Q1 2022 (the peak), burned through that cash by 2024, went back to raise more... and found the door slammed shut. 

Interest rates had spiked, the easy money was gone, and suddenly investors cared about things like "revenue" and "unit economics."

Wild concept, I know.

Pattern #1: The 2021 Overfunding Curse

Here's the uncomfortable truth about raising too much money: Getting overfunded is often worse than being underfunded.

When you raise $50M at a $200M valuation before you've proven product-market fit, you've just locked yourself into a game you probably can't win. 

You need to 10x that valuation for your next round, which means 10x'ing your revenue, which means... you're probably screwed.

Let me show you what this looked like in practice:

Canoo ($1B+ raised): 

EV startup with futuristic designs and massive pre-orders. Filed for bankruptcy January 2025 after disappointing earnings and desperate layoffs. Turns out building cars at scale requires more than slick renderings.

Lilium (eVTOL/flying taxis): 

Years of development, zero commercial flights, declared insolvency in 2025. The technology to make electric vertical takeoff aircraft work commercially just... isn't there yet. No amount of funding changes physics.

Pandion ($125M raised): 

Logistics startup promising to compete with UPS and FedEx. Shut down abruptly in January 2025, laying off 63 employees with basically no warning. Capital-intensive logistics businesses require scale to survive, and they never hit it.

The pattern? 

Companies raised war chests in 2021, spent like drunken sailors, and when it came time to prove they could actually make money, the answer was "lol no."

Underfunded companies are forced to get scrappy, find revenue early, and prove their model works. Overfunded companies hire too fast, spend on "brand building," and convince themselves that scale will solve their unit economics problem.

It won't.

But the most spectacular example of the Overfunding Curse? That belongs to Builder.ai. And their story gets way worse than just raising too much money.

Next week in Part 2, I'm breaking down Builder.ai's $445M implosion—and the fraud that turned "overfunded" into "criminal." The line between ambitious startup and federal crime is thinner than you think.

—Brendan Ward