Paul Graham opened his June 2026 talk with a confession: he'd been astonished. An American politician had just said, publicly, that it was impossible to earn a billion dollars. Not impossible to be a billionaire — obviously that's possible — but that getting that rich without cheating was impossible.
Graham had spent 21 years running Y Combinator, funding roughly 6,500 companies, and watching about 30 founders become billionaires. He felt like a skating coach hearing someone say a triple axel can't be done. Of course it can. It's hard, but it's possible.
The politician was wrong
The argument behind the politician's statement was moral, not mathematical: that extraordinary wealth requires exploitation. That you can't get that rich without taking something from someone. It's an old idea, and it sounds plausible if you've never watched a startup grow.
Graham had just gotten off a call with a founder whose startup was growing at 93% month-over-month. Not 93% per year. Per month. The reason it was growing so fast was simple: users loved what she'd built, and they were telling their friends. No exploitation. The opposite, in fact — relentless work to make customers happy, compounding through word of mouth.
"She was getting richer at a stupendously rapid rate. And yet she hadn't been doing anything bad."Paul Graham, June 2026
Someone replied that having a few million and growing at 93% a month was "radically different from being a billionaire." Graham decided to prove them wrong, live, in front of a room of future heads of state.
The math — do it yourself
Graham asked the room to open their phones and run a calculation. The founder had roughly $2 million in revenue. To become a billionaire, she needed to grow 500×. The question: how many months of 93% growth does that take?
How many months of 93%/mo to grow 500×?
We want log base 1.93 of 500. Type this in Google: log(500, 1.93)
The answer: ≈ 9.45 months
A few million dollars and 93% monthly growth are not radically different from a billion. They are nine and a half months apart.
Nine and a half months. Not radically different — barely a year apart, mathematically. The room understood. This is why growth rate is the only number that matters.
But Graham didn't want to be accused of cherry-picking extreme numbers. So he backed off to something far more conservative.
Only two numbers
What happens at 15% per month? That's not unusual. Graham says he encounters startups growing at 15% a month constantly. It's the benchmark, the rate that marks the boundary between a company that might make it and one that almost certainly will.
What is 1.15 to the power of 60?
Five years is 60 months. Type in Google: 1.15^60
The answer: ≈ 4,384
If you're making $10K/month today, in five years you'll be making $43.8M/month, or roughly $526M/year. At typical founder ownership, you are a billionaire.
| Today's MRR | Year 1 | Year 2 | Year 3 | Year 5 |
|---|---|---|---|---|
| $5K | $26K | $136K | $708K | $21.9M |
| $10K | $54K | $272K | $1.4M | $43.8M |
| $25K | $134K | $679K | $3.4M | $109.6M |
| $50K | $268K | $1.4M | $6.9M | $219.2M |
| $100K | $535K | $2.7M | $13.7M | $438.4M |
The entire model — the entire path to a billion — is captured in two numbers. Graham made this explicit:
"If it's impossible to make a billion dollars without cheating," Graham asked, "which of those two numbers is impossible?" Startups hit 15%/mo without cheating all the time. And market size — how could you possibly cheat to increase that?
Getting the first number
The growth rate comes from one thing: making something users love so much they tell their friends. Graham is blunt about why he always asks founders their growth rate first — it's a proxy for whether they've built the right thing.
You can't fake a good growth rate. You can't acquire your way to 15%/mo in early stage. You can't PR your way there. It's an unambiguous signal of product-market fit. Every month that the number is strong, the market is giving you a vote of confidence. Every month it's weak, it's telling you something is wrong.
The best startup ideas don't come from looking for startup ideas. They come from building things you and your friends want — because you predict future demand. Whatever you and your cohort start using now, the world will be using in ten years. Your own needs are a uniquely valuable signal. Don't override them with market research.
Graham's rule: make something for yourself. You understand yourself. You can't understand other people nearly as well, especially at the beginning. The hack of making something for yourself and your friends has produced Apple, Google, and Facebook. None of them started as companies. They started as things people built because they thought it would be cool.
How growth really decays
Graham acknowledged the real world: growth rates slow. A very successful startup is probably growing faster than 15%/mo in year one and slower than 15% by year four. But he noted something important — you end up in roughly the same place. The compounding still happens. The outcome is still extraordinary.
What does realistic decay look like? Based on what YC has observed across its portfolio, the pattern tends to follow a predictable arc: high early growth that decelerates as the market matures, offset by the company's increasing operational sophistication.
Churn is the invisible tax on this decay. Every point of monthly churn is a permanent drag — it reduces the effective net growth rate and accelerates the point at which growth stalls. A company growing at 8%/mo gross but churning 5%/mo has an effective net rate of only 3%. The compounding is gutted. This is why retention is not a customer success metric — it's a growth metric.
"There are only two numbers in the calculation, the growth rate and how long it continues."Paul Graham, June 2026
What this means for you
Most founders check their revenue occasionally. They look at the dashboard when a new customer comes in, or when something breaks. They don't track growth rate as a primary metric because it requires at least two data points — and thinking carefully about what those data points actually represent.
MRR growth and revenue collected are not the same thing. A company with 100% annual plans can collect $12,000 in one month and nothing the next — but their MRR, the normalized monthly subscription value, barely moves. Tracking collected revenue as a proxy for MRR growth will give you meaningless numbers. An annual invoice hitting in May is not 140% month-over-month growth. It's a timing artifact.
The number Graham cares about is the one that compounds: the change in your Monthly Recurring Revenue from one month to the next. That's the number in the model. That's the number that, sustained at 15%, turns $10K into $43M in five years.
Your growth rate isn't a vanity metric. It's the single most important predictor of your company's future value. The founders who become billionaires are not the ones with the highest current revenue — they're the ones with the highest sustained growth rate. That's the number to watch. Every morning. Every month.
Graham closed with a summary sharp enough to put on a wall: there are two numbers that determine how big a startup gets, and how rich its founders become. The growth rate, and how long it continues. You get the first by making something users love enough to tell their friends. You get the second by being in a big enough market. If you grow exponentially into a big market, the wealth happens automatically — not as a side effect of exploitation, but as a direct consequence of making people's lives better.
The mathematician's answer to the politician's claim is not an argument. It's a calculation. Do it yourself. The answer is 4,384.
Your growth rate,
every morning
RevOS computes your real MRR growth rate daily — not collected revenue, not invoice totals, but the actual change in your subscription base — and tells you exactly where it takes you in 12 months and 5 years. With churn factored in.