Paying Is the Only Signal
Discount friends' feedback to zero, and convert 'willing to pay' into money actually moving — refundable
- Difficulty
- Moderate
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 95%
A two-part validation discipline. First, a hard rule: feedback from friends, well-wishers and anyone who isn't your named target customer is discounted to zero — not weighted down, zeroed, not even written down. Second, replace 'willingness to pay' with an actual payment, offered with a no-questions refund, because the moment money moves the conversation changes and the real requirement surfaces. Weinstein applies the same test to enterprise contracts and partnerships, and even has founders practise charging him a dollar.
Origin
Jeff Weinstein's rule, born from repeatedly falling for friendly feedback himself ('because I've fallen down that path myself and so many times I just set a rule'). He notes Stripe's own founders push the same methodology internally at 8,000+ people when a customer wants Stripe to go off-roadmap. It sharpens the standard lean-startup 'willingness to pay' concept by insisting on the transaction rather than the intent.
Core principles
- 01Your target customer is not a segment or a company size — it's Sarah in this department, who has these tabs open, hit this problem, and needs it solved by 4pm.
- 02People who build things are empathetic and have friends; those friends will click around and say nice things. That is not signal.
- 03A rule beats a weighting, because you (and everyone) are drawn to the friendliness. So: discount friends to zero.
- 04Ready-to-pay, said-they'd-pay and looked-at-the-contract are all categorically different from paid. Paying is an independent person judging their problem burning enough to trade something of value for a promise.
- 05The refund removes all downside from asking, so there is no excuse not to ask.
- 06Attaching a price flips a wishlist ('can it do one, two, three?') into a requirement ('for $10,000 it would need to do X') — and X is the actual value.
How to run it
- 1
Name Sarah
Define your target customer at the level of an individual with a specific job, a specific open tab, and a specific deadline — not 'digital natives above X employees'. Everyone on the team should be able to say her name.
- 2
Zero out everyone who isn't Sarah
Set the rule explicitly: feedback from friends and friendly beta users is of zero interest. It doesn't get written down. It isn't part of what the team discusses at all. Only Sarah's problem counts, and only whether you can solve it as fast and as janky as necessary.
Pro tip Make it a rule rather than a judgement call. 'Pay less attention to it' isn't enough — Weinstein tried that on himself and it failed.
Watch out This will feel harsh and socially costly. That is exactly why it has to be a rule rather than a case-by-case call.
- 3
Put a real price in front of Sarah
Even Sarah will generate a wishlist if you're solving her problem for free. So price it: 'by the way, this is $10,000 — and I'll happily refund you 100% the second you don't like it.' Then listen for what she says it would have to do to be worth that.
Pro tip The pushback is the payload. 'I like this but not for $10,000 — for $10,000 it would need to solve X.' That X is the value you were both dancing around.
- 4
Apply the same test to enterprise and partnership commitments
The siren song of a big multi-year enterprise contract in exchange for building features 1-9 and clearing security steps A-G leads to rugs being pulled mid-build and contracts that never land — the majority case. So ask them to wire the money: 'if we're super serious about this, wire us a million dollars — we'll happily wire it back any time you need it.'
Pro tip Either answer is a win. 'Absolutely not, we can't commit' saves you months of building something never used. 'Sounds great, and now that we're paying we want it faster' puts you on the fast road.
Watch out This applies at any size. There's always a bigger logo — Fortune 1,000 to 500 to 10 to 1 — so the temptation never goes away.
- 5
Make founders practise charging
Weinstein has founders send him a real payment link or invoice for one dollar, right there on the call — any payment provider, he doesn't care which. So that when it's time to charge their first real customer, it isn't their first time, the logo is already on the invoice, and there's already a dollar in the account.
Pro tip He says his inbox is full of $1 receipts to random people. The point is to eliminate the awkwardness of the first charge before the moment that matters.
In the wild
Weinstein describes the pattern: because you're attentively solving Sarah's problem, she naturally says 'this is great, I want features X, Y and Z, can it do one, two and three?' Then you say the thing costs $10,000, fully refundable the second she doesn't like it. She stops: 'whoa, wait a second — I like this thing, but not for $10,000. For $10,000 it would need to solve X.'
→ 'Oh, there it is' — the price reveals the single capability that carries the value, which no amount of free feedback would have surfaced.
On calls with founders, Weinstein asks them to sign up for any invoicing or payment service (he doesn't care which, though he's happy to hear feedback if they pick Stripe) and send him a payment link or invoice for one dollar immediately, on the call.
→ The founder has now charged someone, has their logo on the invoice, and has a dollar in their account — so the first real customer charge isn't a novel, weird act.
Common mistakes
Counting friendly beta feedback
Friends will use your beta, click around and say they liked it — because they know you and like you. That is extremely different from Sarah, who has the actual problem and is willing to pay for it. The pull toward the friendly group is strong enough that it needs a hard rule.
Treating 'ready to pay' as validation
Reviewing the contract, saying they'd pay, and even signing intent are all categorically different from money moving. Weinstein calls the difference night and day, and says he too used to feel good about the contract stage.
Building the enterprise feature list before the wire
The nine-feature, seven-security-step, three-year-contract deal is the classic rug-pull. The majority case Weinstein hears is that the contract never lands and the software is never used — a huge amount of build time destroyed.
Is it for you?
Best for
B2B founders and 0-to-1 teams deciding whether a problem is real enough to build against, and anyone weighing a large off-roadmap enterprise or partnership commitment
Not ideal for
Consumer and social products with indirect monetisation, where Weinstein explicitly says he has less intuition, and free/internal tools with no price surface
From the transcript
“not digital natives that are X big I'm talking about Sarah in this department who has these tabs open and just faced this problem and…”
“that is not of interest to us it doesn't we don't even write it down it's not part it's just not part of what we're…”
“if you said and by the way this thing is $10,000 um Phil will happily refund your money 100% the second you don't like it”
“independent group of people saying my problem is burning enough that I'm willing to exchange something I have that has value for the promise of…”
“if we're super serious about this send us a million dollars like w wire us a million dollars we'll happily wire it back anytime you…”
“send me an invoice or a payment link for a dollar right now like right now”
From the episode
Building product at Stripe: craft, metrics, and customer obsession
Jeff Weinstein (Product lead)