Default Alive or Default Dead: Paul Graham's Startup Survival Test Explained

The First Question Paul Graham Asks

In October 2015, Paul Graham published a short essay with a blunt title: "Default Alive or Default Dead?" The whole idea fits in one sentence. When Graham meets a startup that has been running for a while, the first thing he wants to know is this: "Assuming their expenses remain constant and their revenue growth is what it has been over the last several months, do they make it to profitability on the money they have left?"

If the answer is yes, the company is default alive. Left alone, with no new funding and no heroic changes, it drifts toward profitability and survives. If the answer is no, the company is default dead. On its current path it runs out of cash and dies unless the founders raise more money or change something big.

The word "default" is doing the real work. It describes where you end up if nothing changes. Most founders think about the upside case, the version where the next big customer lands or the next round closes. Graham's question strips all of that away and asks what happens on autopilot. That's a colder and far more useful thing to know.

The framing matters because it turns a fuzzy anxiety into a testable fact. "Are we going to be okay?" is a feeling. "Do we reach profitability on the cash we have?" is a calculation you can run this afternoon. This is the same shift that separates careful founders from hopeful ones, and it's the discipline that keeps a company solvent long enough to matter.

How to Run the Calculation

You need three numbers, and every startup already has them.

1. Cash on hand: how much money is in the bank right now.
2. Monthly burn: expenses minus revenue, the net amount you lose each month.
3. Revenue growth rate: how fast revenue has grown over the last several months, not the last week and not your forecast.

The trick is to project forward honestly. Hold expenses flat. Grow revenue at the rate you've actually been hitting. Then watch the two lines. Every month, revenue rises, so your net burn shrinks. If revenue catches up to expenses before the bank hits zero, you're default alive. If the bank hits zero first, you're default dead.

Here's the same test in a table you can fill in yourself:

The honesty of the inputs is everything. A founder who assumes 20% monthly growth because one good month happened, or who leaves out the five hires planned for next quarter, will get a comforting answer that means nothing. Run the pessimistic version too. If you're default alive only under your best-case growth, you're really default dead with a nice story attached.

Reading founder essays closely helps here, because the good ones show their real math. When you study pieces like this, save the exact numbers and definitions rather than the vibe. Glasp's web highlighter lets you highlight the key passages of essays like Graham's straight from the page, and your highlights stay searchable so the framework is there the day you actually need to run the test.

Why Half of Founders Don't Know

The most quoted line in the essay is also the most alarming. "Half the founders I talk to don't know whether they're default alive or default dead," Graham writes. These aren't careless people. They're smart, driven founders running real companies. They simply haven't done the arithmetic, or they've done it with numbers soft enough to avoid the answer.

Not knowing is not neutral. It's the actual danger. A default dead company that knows it's default dead can act: cut costs, push revenue, raise while it still looks strong. A default dead company that thinks it's fine keeps spending as if the money will last, and discovers the truth only when the runway is too short to fix anything.

Graham's advice on timing is deliberately aggressive. "Instead of starting to ask too late whether you're default alive or default dead," he writes, "start asking too early." Early is uncomfortable because the answer might force hard choices while things still feel okay. That discomfort is the point. The question is cheap when you have eighteen months of runway and brutal when you have three.

This is a habit, not a one-time exam. The number moves every month as you spend, hire, and grow. Founders who track it treat it like a vital sign, checked constantly, not a checkup done once a year. The best ones fold it into how they run the whole company, watched as routinely as revenue or headcount.

The Fatal Pinch

Graham named the worst version of default dead the fatal pinch. His definition is precise: "The fatal pinch is default dead + slow growth + not enough time to fix it." Three conditions stack, and together they close the exits.

Walk through why each piece hurts. Default dead means the current path ends in zero. Slow growth means you can't outrun the burn by selling your way out. Not enough time means the runway is too short to change the trajectory before the cash is gone. Any one of these is survivable. All three at once is the pinch.

What makes it fatal is a psychological trap on top of the financial one. Founders in the pinch tend to overestimate their odds of raising more money. So instead of cutting costs or chasing profitability, they bet on a round that probably won't come. The waiting burns the runway they had left, and a company that isn't growing and is running low on cash is exactly what investors avoid. The delay that felt like patience is what seals the outcome.

The escape is to act while you still have options. A default dead company with a year of runway has real moves: it can cut, it can push revenue, or it can raise from a position that still looks like strength. The same company with two months of runway has almost none. Time is the resource that converts a bad diagnosis into a survivable one, and it's the one founders waste by waiting to look.

Hiring Is the Biggest Killer

Ask a founder in trouble what went wrong and you'll rarely hear "we hired too fast." Graham says that's usually the answer anyway. "Hiring too fast is by far the biggest killer of startups that raise money," he writes. The logic is subtle. Founders see that big successful companies have lots of employees, so they reason that hiring will make them successful. They have cause and effect backwards. The headcount is a result of growth, not the cause of it.

His counterexample is Airbnb. The company waited four months after raising money at the end of Y Combinator before hiring its first employee. The founders were terribly overworked in that stretch, but they spent the time refining the product instead of padding the org chart. Staying small kept them close to the problem and kept the burn low, which is exactly what a young company needs.

The 2022 correction turned this warning into a public confession from some of the strongest companies in tech. On November 3, 2022, Stripe cut roughly 14% of its staff, about 1,100 people. CEO Patrick Collison's email to employees didn't hide behind euphemisms: "We overhired for the world we're in." He named two specific errors. The company was "much too optimistic about the internet economy's near-term growth in 2022 and 2023," and it "grew operating costs too quickly." Stripe was not a failing company. It was a healthy one admitting, in plain words, the exact mechanism Graham warned about.

Others made the same admission. Coinbase's CEO acknowledged the company had over-hired during the boom. Rapid-delivery startups that had staffed up for endless growth, like Getir, cut thousands of roles when the funding turned. The pattern repeats because the temptation is structural. Money in the bank feels like permission to spend, and the easiest thing to spend it on is people, who are also the hardest thing to unwind.

The 2022 Reckoning

For years, cheap capital let founders ignore Graham's question. If you could always raise more, "default dead" was just a phase between rounds. That assumption broke in 2022, and the break was fast.

In May 2022, the warnings arrived almost at once. Y Combinator sent portfolio founders a letter that dropped the usual optimism. "No one can predict how bad the economy will get," it read, "but things don't look good." The advice was to "plan for the worst," to cut costs, and to extend runway. It was pointed about fundraising: if your plan was to raise in the next six to twelve months, you might be raising into the worst of the downturn, so you should change the plan. Days earlier, Sequoia had presented a fifty-plus slide deck titled "Adapting to Endure" to its founders, calling it a "Crucible Moment" and warning that the market was no longer rewarding growth at any cost. The companion advice was blunt in the same direction: extend your runway, become a real business.

Both messages amounted to the same instruction Graham had written seven years earlier. Stop assuming the next round. Get to a place where you survive on your own. The vocabulary was runway and profitability, but the target was default alive.

The companies that couldn't get there showed up in the data. According to Carta, roughly 770 US venture-backed startups shut down in 2023, up sharply from 467 in 2022, and the number rose again to 966 in 2024, a record. Layoffs told the same story. Per layoffs.fyi, tech companies cut about 164,000 workers in 2022 and roughly 263,000 in 2023. A generation of founders learned Graham's lesson the expensive way: the default matters most exactly when you can no longer paper over it with a new round.

Default Dead in the AI Era

The current wave of AI startups makes the question harder, not easier, because it hides behind the most impressive growth charts in years. Revenue can rocket from nothing to hundreds of millions in months. That looks like the opposite of default dead. Sometimes it isn't.

The reason is unit economics. Classic software companies enjoy gross margins of 70% to 80%, because serving one more customer costs almost nothing. AI products don't work that way. Every query runs on expensive compute, so the cost of serving customers scales up right alongside revenue. The dominant cost is inference, and it grows when you succeed.

The clearest reported example is Anthropic. According to The Information, which reviewed internal figures, Anthropic's 2024 gross margins ran roughly negative, somewhere around minus 94% counting only paying customers and worse counting everything, meaning it cost more to serve usage than the usage brought in. Inference costs came in about 23% above projections. The company reportedly cut its 2025 gross-margin target to around 40%, even as revenue scaled into the billions. Anthropic hasn't confirmed these figures publicly, so treat them as reporting rather than gospel. The lesson holds either way: a company can grow revenue explosively and still be losing money on every unit of what it sells.

Run Graham's test on a business like that and the growth line is a trap. Faster growth means faster burn, not faster profitability, until the margins turn. A startup can be default dead precisely because it's growing, if each new customer widens the loss. Sorting real durability from a good-looking chart is the whole skill, and it's the same discernment covered in product-market fit and in competition is for losers. Growth without margin is a countdown with a nicer soundtrack.

The Gumroad Turnaround

The most useful part of Graham's framework isn't the diagnosis. It's that default dead is often reversible if you catch it in time. The cleanest real example belongs to Sahil Lavingia, founder of Gumroad, who wrote about it with unusual honesty in his 2019 essay "Reflecting on My Failure to Build a Billion-Dollar Company."

Gumroad had raised about $8.1 million, including a $7 million Series B led by Kleiner Perkins in 2012. The plan was hypergrowth. It didn't come, a follow-on round fell through, and the company was default dead: burning far more than it earned with no rescue in sight. In early 2015, Lavingia made the cut most founders dread. He laid off 75% of the team, going from 20 employees down to 5.

The numbers show what that bought. In June 2015, before the deepest cuts fully took hold, Gumroad was making about $89,000 a month in revenue against roughly $364,000 in operating expenses, losing around $351,000 every month. A year later, in June 2016, revenue had grown to about $176,000 a month while operating expenses had been slashed to around $32,000. The company netted a small profit, roughly $10,000 a month. It had crossed from default dead to default alive.

The story isn't a fairy tale. Lavingia is candid that Gumroad never became the billion-dollar company its funding implied, and getting there hurt. But the company survived, kept serving creators, and years later was still operating and growing. That's the underrated payoff of Graham's question. Being default alive isn't glamorous, and it won't make headlines, yet it's the state that lets a company keep existing long enough to matter.

The founders who make these calls well tend to be voracious, careful readers of other people's hard-won lessons. Graham's essays, Lavingia's postmortem, the Stripe memo: these are primary documents, and studying them beats absorbing a diluted summary. You can summarize founder talks and YC videos on YouTube to pull timestamps and key points fast, then ask Glasp's AI chat questions across everything you've saved, so the next time you run your own number the whole playbook is at hand. It's the same practice behind do things that don't scale and how to get startup ideas: learn from the primary sources, in public, and keep what you learn.

Frequently Asked Questions

What does default alive or default dead actually mean?

It's Paul Graham's test for whether a startup survives on its own. Assume your expenses stay flat and your revenue keeps growing at the rate it has recently. If you reach profitability before your cash runs out, you're default alive. If you run out of money first, you're default dead. The word "default" means the outcome if nothing changes, no new funding and no dramatic pivot.

How do I calculate whether my startup is default alive?

Take three numbers: cash in the bank, your monthly net burn (expenses minus revenue), and your real trailing revenue growth rate. Project forward with expenses held flat and revenue growing at that rate. Each month your net burn shrinks as revenue rises. If revenue catches expenses before the bank hits zero, you're default alive. Use honest inputs. Don't count unclosed funding, and don't plug in growth you hope for instead of growth you've hit.

What is the fatal pinch?

Graham defines it as "default dead + slow growth + not enough time to fix it." All three at once is what makes it deadly. Founders in the pinch usually overestimate their chances of raising more, so they wait instead of cutting costs, and the delay makes them even less fundable. The escape is to act early, while you still have runway and options.

Is it bad to be default dead?

Not necessarily, if you know it and you're early. Many great companies were default dead at some point. The danger is not knowing, or knowing too late to change course. A default dead company with a year of runway can cut costs, push revenue, or raise from strength. The same company with two months left has almost no moves. Default dead is a warning to act, not a death sentence.

Why is hiring the biggest killer of funded startups?

Because founders treat headcount as a cause of growth when it's actually a result of it. Raising money makes hiring feel safe, so companies staff up ahead of real demand, and people are the hardest cost to reverse. Airbnb waited four months after its raise to hire anyone. In 2022, Stripe cut about 1,100 roles and its CEO said flatly, "We overhired for the world we're in."

Does fast revenue growth mean I'm default alive?

No. Growth and default-alive are different things. If your gross margins are negative, faster growth means faster losses, not faster profitability. Some AI companies have grown revenue into the billions while reportedly losing money on every unit served, because inference costs scale with usage. Run the actual projection. Growth only makes you default alive if each new customer moves you toward profit, not deeper into the hole.

Conclusion: Know Your Number

Paul Graham's question survives because it's simple, honest, and impossible to fake once you've run it. Default alive or default dead is not a mood or a pitch. It's a number, and it's sitting in your bank statement and your growth rate right now, waiting to be calculated.

The founders who thrive don't necessarily start out default alive. They start out knowing which one they are. That knowledge is what turns a bad trajectory into a fixable one, because it buys the thing the fatal pinch steals: time to act. Gumroad found that time by cutting hard and early. Thousands of companies in 2022 and 2023 found out too late.

So run the number. Then run it again next month. Highlight the essays and postmortems that sharpen your judgment, from Graham to Lavingia to the memos of founders who learned in public, and keep them somewhere you can search. Save the passages that matter and use the community feed to see what other builders are marking up in the same essays. The test takes an afternoon to learn and a lifetime to keep passing. Start asking too early. That's the whole advice.