All about subscription #3
A few weeks after we launched our second product, I pulled up the revenue chart and looked for the slide.
I knew the slide well. It was the shape of my first product: every campaign a spike, every quiet week a slide back toward zero. I wrote about it in the last post — we hadn't built a money machine, we'd built a campfire, and I had spent [a year] as its full-time wood gatherer.
The marketing push for the new product had ended. The slide should have started.
It wasn't there. The line just... stayed.
Recurring revenue's first gift isn't growth. It's sleep.
That's my whole argument. Growth came later. Sleep came the first month.
LOVE to a code
The new product was a dating service, over text messages, in 2003.
You texted LOVE to our service code. The service asked you a short survey — basic personal information, nothing clever. Then it matched you with someone and the two of you chatted, anonymously, through us. You stayed anonymous until you both felt a connection; then you took it to real phone numbers, and real life. If a match didn't feel right, you walked away, no awkwardness, and the service never paired the two of you again.
The price was three dollars a month. Unlimited chat. Unlimited matches.
Millions of young, lonely people joined and paid the three dollars without hesitation. Nothing else at that price did anything like it.
Who texts a stranger at midnight
To see why the price had to be flat, look at who was paying it.
The confident ones didn't need us. They had friends of friends, parties, coworkers, nerve. The people who texted LOVE to 3030 were the other kind: young, shy, short on money, and long on feeling. For them, the phone was the most private space they owned — no screen anyone could look over, no profile anyone could stumble onto. And texting was already the native language of flirting. We didn't invent a behavior. We gave an existing one somewhere to go.
A shy person walks into a dating service with the first fear of being judged — showing your face, saying the wrong thing, being rejected out loud. Product design answered that one: anonymous until you both chose otherwise, walk away anytime, never rematched with someone you'd left. Rejection was engineered down to a shrug.
The second fear is quieter, and pricing is the only thing that can answer it: the meter.
The meter kills the mood
Put a price on each message and you put a price sticker on every sentence. A shy person drafting something honest at 2 a.m. does not need one more reason not to hit send. Metered pricing makes people ration — and rationing is fatal to exactly the kind of conversation the product existed to create: tentative, open-ended, a little longer than it strictly needs to be. That's not waste. That's the raw material of falling in love. At ten cents a text, we would have been taxing it.
Underneath that sat a colder fear: the bill. Our customers had thin wallets. Under per-text pricing, the month the product finally works — the long conversations, the late nights — is the month they can't afford. Nobody opens their heart on a taxi meter.
Now notice the shape of the problem. Per-use pricing gave us volatile revenue and gave them volatile cost. The same uncertainty, losing sleep on both sides of the transaction. Three dollars flat retired it for both of us at once: they knew their worst case, we knew our base case. Chat all night. It costs what it costs.
And the flat rate did one more thing I only fully appreciated years later. At ten cents a text, our best customer is the one who talks too much — the pricing pays us to prolong conversations. At three dollars a month, our best customer is the one who renews — the pricing pays us to make matching actually work. Your price model isn't just what you charge. It's a public announcement of what you're optimizing for, and customers can smell it.
From campfire to furnace
In the last post I said our per-text business was a campfire: warm while you feed it, ash the morning after. The dating service taught me how the metaphor ends.
A subscription is a furnace.
Same fire, different relationship to it. A campfire needs someone gathering wood every single day, and the house goes cold the day you stop. A furnace runs on a fuel contract: heat arrives on schedule, and nobody's job is wood. The fire didn't get bigger when we switched — at three dollars a head, it arguably got smaller. It got reliable. And reliable changes what you're allowed to build around it.
Around a campfire, you camp. Around a furnace, you build a house. We could plan next quarter because we could see next quarter. We could hire because we knew payroll would clear in month four. We could spend our days improving matching instead of gathering wood — which, translated out of metaphor, means the marketing budget finally became product budget.
Within twenty-four months of founding, the company was acquired. I don't think that happens to the campfire version of us. Nobody acquires a fire that goes out when you stop feeding it. They acquire a furnace with a fuel contract.
The honest part, again
The pricing switch gets too much credit when I tell this story quickly.
The deeper switch was the product underneath it. Trivia is bursty — you binge it, and you're done. The need for connection renews. And it doesn't just renew — it circles. Customers left us when the chatting turned into real dates. Then the dates didn't work out, or the relationship did and then didn't, and they came back and texted LOVE again. Our churn had a return path built into it by human nature.
I should also be honest about what I was rooting for. It would make a lovely story to say our happiest moment was watching a customer walk away matched, done with us forever. It would not be a true story. I'm a businessman, not a matchmaker. The only exit we never won back was the customer who found a long-term relationship — the one competitor we could never out-price was marriage. I could live with losing to that. I did not root for it.
So no, you can't bolt a monthly price onto any product and buy yourself sleep. The subscription didn't create the demand. It revealed a demand that was already shaped like a month — and, in our case, shaped like a circle. Part of the science — most of it, maybe — is telling the difference.
What three dollars actually bought
The customers got matches. We got something stranger: mornings that didn't start from ash.
But the furnace comes with a condition the campfire never imposed. Nobody re-decides a log after it's burned — that money already happened. A fuel contract gets re-decided every single month, by every single subscriber, forever.
That condition is not a footnote. It's the whole discipline, and it's the next post.
Until then: look at your own product and ask one question. Is the reason people pay you shaped like a moment, or shaped like a month? Don't answer with your pricing page. Answer with their behavior.
Where does this argument break?