TLDR
- In September I spent weeks sharpening a framework and collected five objections and zero buyers. Refining the idea was a way of avoiding the question.
- The rule since: nothing gets built until ten people have said yes to a number. Not to the idea. To the price.
- A yes to an idea is free. “I’d pay that” is the only market research that counts, and “I bet it’s great, I just don’t have time” is a no.
- Second rule: if an AI can answer it competently, it can’t be sold. The thing has to produce something a chatbot can’t: your number.
- I’m running it on this site now, in the open. Two decision sheets, one for each door. Ten names each. The one that reaches ten first gets built first.
Five objections. Zero buyers. That was September’s scorecard.
I’d been working on a framework, the kind of thing you can draw on a napkin and feel clever about. Every week I made it sharper. Every week I found another objection and answered it. By the end of the month it was airtight, and I had not asked a single person whether they would pay for it.
That’s not diligence. That’s hiding. Refining the thing is a way of never finding out whether the thing is wanted, because as long as it isn’t finished, nobody can say no to it.
The rule
Since then, one rule, and I took it from Ramit Sethi because he’d tested it harder than I ever will: nothing gets built until ten people have said yes to a number.
Not yes to the idea….cos everyone says yes to an idea – it costs nothing and it’s polite. Yes to a price: “If I build this and it does what I just described, it’ll be this much. Does that work for you?”
Ten of those and the launch is close to certain. Fewer than ten and the offer is wrong, or the price is, or the person is, and I’d rather learn that in a phone call than after three months of building.
How to hear the answer
People are kind, so the answers come disguised. Three kinds:
- A no that sounds like a yes. “I bet it’s great, I just don’t have time right now.” Translation: I don’t care enough. Count it as a no and thank them.
- A real yes. “That’s a bit high, but it’s fair, and I’d seriously consider it.” This is the one most people throw away because it isn’t enthusiastic. It’s the best data you’ll get.
- The rare one. “Take my money.” Lovely. Also the least informative, because that person would have bought anything you made.
When the answer is vague, two questions: “What’s holding you back?” and “What would make this a no-brainer?” Then stop talking.
The second rule
There’s a filter before the ten calls, and it’s new this year. If an AI can answer the question competently, it can’t be sold.
Most of what I know about running a service business is now a chatbot’s answer away. Hiring, pricing, leases, the lot. Writing it down and charging for it is selling bottled water next to a tap. What a chatbot can’t do is produce your number: what your business is worth without you in it, what your date is if you take one step up the ladder, what the first client outside your job does to the maths. That needs your inputs and a method, and the method is the product.
If it doesn’t produce a number the reader couldn’t get elsewhere, it’s a blog post. Give it away.
Running it in the open
So here’s what I’m doing with this site, and I’d rather you watched than guessed.
Two decision sheets, one for each door on the calculator. For the reader with no business yet: the first paying client outside the job, the move that took me from $1,700 a month to a net $4,800. For the owner: what the business is worth with you in it and without you, and the three moves that close the gap, each priced in years.
Ten names for each. I’ll ask every one of them for a yes to a number. Whichever list reaches ten first gets built first. The other follows. If neither gets past five, the sheets are wrong, not the readers, and I’ll say so here.
If you’d be one of the ten, on either door, reply to Friday’s letter and tell me which. I’ll ask you the question, and I’ll take the honest answer over the kind one.