The YC Graveyard: What 800+ Failed Y Combinator Startups Teach Us About Building
The YC Graveyard: What 800+ Failed Y Combinator Startups Teach Us About Building
TL;DR: The YC Graveyard now lists over 800 inactive Y Combinator startups. From NFT exchanges to factory AI — the reasons vary, but the patterns repeat. Here's what the graveyard teaches us, with real examples and data from the YC class of Winter 2022 alone.
The Numbers
Y Combinator has funded over 5,000 startups since 2005. Roughly 60% are still active. That leaves about 40% that have either died quietly, been acqui-hired, or faded into inactivity.
The YC Graveyard — a project by Will Wang that tracks inactive YC companies — has documented 821 dead startups as of May 2026. That's a lot of dead companies backed by the world's most prestigious accelerator.
But here's the surprising part: YC's failure rate (~18-20%) is dramatically lower than the startup average (~90%). Even the dead YC companies tend to last longer, raise more money, and go further than non-YC startups.
The graveyard isn't a list of failures. It's a list of learning opportunities.
Winter 2022: A Case Study in Timing
The Winter 2022 batch is particularly interesting. These companies launched into a unique moment in tech history — and many didn't survive.
Here are some of the inactive companies from that batch and what they tried to build:
NFTScoring — The NFT Exchange for Traders
One-liner: "The NFT exchange for traders"
Team size: 8
Industry: Consumer / Crypto
Location: New York
Launched in January 2022 at the absolute peak of NFT mania. By the time they'd built their product and found product-market fit, the floor had dropped out of the NFT market entirely. Crypto winter hit in mid-2022, and an exchange for NFT traders became a solution in search of a problem.
Lesson: Timing isn't everything — but it's close. If your business depends on a frothy market staying frothy, you're building on sand.
Cashmere — Crypto Wallet for Web3 Startups
One-liner: "Crypto wallet for web3 startups"
Team size: 3
Industry: Fintech / Crypto
Location: San Francisco
A 3-person team building a Solana multisig wallet for enterprises. The idea was solid — startups needed better wallet infrastructure. But 2022-2023 was brutal for crypto infrastructure plays. FTX collapsed. Solana nearly died. Enterprise demand for crypto wallets evaporated.
Lesson: Small teams can move fast, but they're also fragile. One macro shock can wipe out your entire addressable market if you don't have the runway to wait it out.
AiSupervision — The Operating System for Factory Production Lines
One-liner: "The operating system for factory production lines"
Team size: 13
Industry: Industrials / Manufacturing
Location: Mannheim, Germany
This one hurts. 13 people, a clear problem (factories need better production tracking), and a solid B2B value prop. They automate what the best supervisors would do — watching everything happening on the factory floor.
The challenge with industrial B2B: enterprise sales cycles are 12-18 months and factories move slowly. Even with a great product, you run out of money before the contracts close.
Lesson: B2B industrial is a marathon. Make sure your runway matches your sales cycle.
Flike — AI Sales Emails That Convert
One-liner: "Flike's AI crafts sales emails that convert"
Team size: 8
Industry: B2B / AI
Location: San Francisco
An AI sales email platform that trained models on your product marketing knowledge. They were even "partnering with some of the largest YC companies." But in 2022, the AI sales assistant space was already crowded — and then ChatGPT launched.
The irony: Flike was doing AI sales emails before it was cool. But by the time the market was ready, everyone was doing it, and doing it with foundation models that made Flike's custom training approach look expensive and slow.
Lesson: Being too early is indistinguishable from being wrong — until it isn't. But if you're too early AND capital-intensive, you usually don't survive long enough to be right.
Vecindario — Real Estate + Mortgages in LATAM
One-liner: "Make it easier for real estate to sell properties and mortgages"
Team size: 87
Industry: Real Estate / Fintech
Location: Medellín, Colombia
An 87-person team at "Growth" stage — this was a real company that had real traction. They were speeding up home buying and mortgage approval in Latin America by 6x. And they still went inactive.
LATAM proptech is a capital-intensive business. Rising interest rates in 2022-2023 made mortgages more expensive across the region. When the macro environment turns against you in an emerging market, there's often nowhere to hide.
Lesson: Scale doesn't guarantee survival. If your business model is sensitive to interest rates and you're operating in emerging markets, you need either massive margins or massive reserves.
The Patterns Behind the Graveyard
Looking across the 800+ inactive YC startups, five patterns emerge:
1. Market Timing Mismatch
The single biggest killer. Companies that launched during crypto mania or the ZIRP (zero interest rate policy) era and were optimized for a world that no longer exists. When the macro shifts, business models that worked at zero percent interest stop working at five percent.
2. Premature Scaling
Multiple teams of 8-13 people building products that didn't yet have proven demand. YC encourages "do things that don't scale" — but some teams scaled their headcount and burn rate before they scaled their revenue.
3. Enterprise Sales Velocity
B2B and industrial startups that had real products solving real problems — but couldn't close deals fast enough. Enterprise sales cycles killed them, even when the product was good. If your smallest contract takes 12 months from first call to signed deal, you need 24 months of runway minimum.
4. Capital-Intensive Business Models
Real estate, manufacturing, hardware — these require capital to operate. When venture capital dried up in 2022-2023, capital-intensive startups starved first. The market rewarded software margins, not asset-heavy models.
5. Single Point of Failure
Small teams (3-8 people) with concentrated expertise or a single customer channel. One co-founder leaving, one key customer churning, one regulatory change — and the company folds.
What YC Startups Get Wrong (According to Paul Graham)
Paul Graham, YC's co-founder, famously outlined 18 mistakes that kill startups. The top three:
- Not making something users want — still the #1 killer, even for YC companies
- Bad founder relationships — about 20% of YC startups lose founders to disputes
- Perpetual side projects — founders who don't commit fully rarely succeed
The YC Graveyard validates all three. Many of the dead companies had interesting technology, good teams, and decent execution — but they were solving problems users didn't care enough about.
Why This Matters
The YC Graveyard isn't about schadenfreude. It's a dataset that's unusually transparent for the startup world.
Most startup failures are invisible — companies just quietly shut down, delete their websites, and disappear. YC companies, by contrast, are often public about their journeys. They leave behind company pages, blog posts, and sometimes even post-mortems.
This transparency makes the YC Graveyard one of the best resources in the world for understanding what actually kills startups. Not theories. Not VC platitudes. Just 821 real companies that tried, almost succeeded, and didn't.
Before You Build
If you're starting a company in 2026, here's what the graveyard says:
- Check your timing. Is the market growing or shrinking? If you're betting on a recovery, make sure your runway is long enough to see it.
- Keep your burn low. The teams that survive the longest are the ones that need the least revenue to break even. Every dollar of monthly burn is a month of optionality you're selling.
- Talk to users until it hurts. If you're building something and people aren't excited, it doesn't matter how good the tech is. The graveyard is full of well-engineered products nobody wanted enough to pay for.
- Build for the market that exists. The market that existed in 2021 is never coming back. The market that will exist in 2028 doesn't matter yet. Build for right now.
The YC Graveyard is a sobering place. But if you read it carefully, it's also one of the best startup education tools available — you just have to be willing to learn from other people's failures instead of your own.
Check out the YC Graveyard by Will Wang for the full list.
Tags: News, Guides, AI, Productivity
Data source: YC Graveyard, Y Combinator public company database
← Back to all posts
