How to measure product market fit: count repeat buyers

How to measure product market fit in a brand that already sells: count the customers who come back without being chased. The first number is the repeat share, orders minus unique customers divided by orders, and three more complete it: the 90-day repurchase rate, repeat purchase by launch cohort and the payback of a new customer. First orders prove problem-solution fit; strangers who come back prove product-market fit.

Most creative brands reach that point the same way. The founder launches, the first drops sell through to friends, followers and early fans, and the brand grows on that momentum. Then the question changes. Selling out once says the idea works. It does not say whether there is a market beyond the people who already knew the founder, and that gap is what the four numbers measure.

To measure product market fit in a brand that already sells, count repeat buyers: the repeat share of orders, the 90-day repurchase rate, repeat purchase by launch cohort and the payback period of a new customer. First orders prove problem-solution fit; strangers who come back on their own prove product-market fit.

Key takeaways

• First orders prove problem-solution fit; repeat buyers from outside the first circle prove product-market fit.
• Four numbers measure it: repeat share of orders, 90-day repurchase rate, repeat purchase by launch cohort and payback period.
• Sean Ellis's survey adds a qualitative check: 40 % or more 'very disappointed' answers.
• Repeat share equals orders minus unique customers, divided by orders, read separately for the first circle and everyone else.

Problem-solution fit vs product market fit: two different proofs

Strategyzer separates the two stages. At problem-solution fit, you have evidence that customers care about certain jobs, pains and gains, and a value proposition designed to address them. At product-market fit, you have evidence that the value proposition is actually creating value for customers, and the offer is gaining traction in the market. A third stage, business model fit, adds the proof that all of it sits inside a profitable model that can scale.

Marc Andreessen's 2007 definition is shorter: product/market fit means being in a good market with a product that can satisfy that market. When it is happening, he writes, customers are buying the product just as fast as you can make it.

For a creative brand, the line between the two proofs runs through the first circle:

• Problem-solution fit is the first launches selling through to people who already follow the founder. It proves the problem is real and that people will pay for this answer to it.
• Product-market fit is people outside that circle buying, and coming back on their own. It proves the brand has a market, not only an audience.

A sold-out first drop belongs to the first list. It is good news, and it is not yet the second proof.

Why first orders can mislead

First orders carry three kinds of noise. Some buyers order to support the founder, and would have ordered almost anything. Some order because a drop was scarce and the date was announced, which measures the launch, not the product. And some arrive once from a campaign and never return. In a revenue chart, all three look the same.

What separates them is the second order. A customer who comes back months later, without a discount and without a personal message from the founder, is the cleanest signal a store has that the product, not the launch, did the work. That is why measuring product market fit in a store is mostly measuring repeat buyers, and why the reading is worth splitting between customers who came from the first circle and customers who found the brand some other way.

How to measure product market fit: four numbers

Shopify's guide to product-market fit lists products selling out consistently and high repeat purchase rates among the signs to look for. Consistently is the operative word: one sold-out drop is a launch, a pattern of returning buyers is a market. Each of the four numbers below comes out of the store admin and an export of orders, and none of them needs a new tool.

1. Repeat share of orders. Orders placed by customers who had already bought, divided by all orders in the last twelve months. It says how much of the business already rests on recurrence. Over a twelve-month window, a close approximation is:

Repeat share = (orders − unique customers) ÷ orders

2. 90-day repurchase rate. Of the customers who placed a first order in a given month, the share who placed a second one within 90 days. The repeat purchase rate guide prices one point of repeat purchase on the blog's worked store.

3. Repeat purchase by launch cohort. Group first-time buyers by the launch that brought them in and follow each group month by month, which is what a SaaS would call cohort retention. If each new cohort comes back at a similar or better rate than the one before, the brand is finding fit beyond the first list. If each cohort fades faster, the launches are buying attention, not customers.

4. Payback period. The months a new customer takes to give back, in gross margin, what it cost to acquire. On the worked store a customer leaves about €49 of margin in twelve months (€68 × 1.6 orders × 45 % margin), roughly €4 a month. Paying the €16 per customer that a 3:1 customer lifetime value target allows comes back in about four months.

Next to the four numbers, a survey adds the qualitative check. Sean Ellis's question asks people how they would feel if they could no longer use the product: very disappointed, somewhat disappointed or not disappointed. After benchmarking nearly a hundred startups, Ellis found that companies with strong traction almost always passed 40 % of 'very disappointed' answers, and companies that struggled to grow almost always fell below it. Superhuman's first reading was 22 %. For a store, send it to customers with at least one order, and split the answers between the first circle and everyone else.

Moving the numbers, launch by launch

Measuring is half the job. The other half is deciding what each launch should test, so the four numbers move for a reason you can name. That is the experimentation system behind build measure learn: every launch carries one hypothesis about one of the numbers, a threshold written in advance and a decision recorded when it closes.

Three loops tend to matter first in a brand that sits between the two proofs:

• Reach beyond the first circle. A limited paid test to an audience that has never heard of the founder, read by how many of those buyers order again within 90 days, not by the first order.
• Earn the second order. One message to first-time buyers around day 30, measured by the 90-day repurchase rate of that cohort against the previous one.
• Tag every launch. Record which launch brought each first-time buyer, so the cohort curve exists before the next launch, not after it.

These are marketing experiments that fit a launch calendar, and each one targets one of the five moments of growth marketing for creative businesses: reaching strangers is Appear, the second order is Retain, and the launch record is Know.

Your repeat share, in one number

Start with the two counts the admin already holds: the orders of the last twelve months and the unique customers who placed them.

Repeat share = (orders − unique customers) ÷ orders
Value of one point of repurchase = customers × 1 % × average order

On the worked store (300,000 sessions a year, a €68 average order, a 1.0 % conversion rate), 3,000 orders come from 1,875 customers at 1.6 orders each. That leaves 1,125 repeat orders: 37.5 % of orders and €76,500 of the €204,000 in revenue. The same store has a repeat purchase rate of 40 %, because 750 of the 1,875 customers ordered again, 2.5 orders each: the rate counts people, the repeat share counts orders. Every point of repurchase among those 1,875 customers is 18.75 second orders, or €1,275 a year.

The number becomes a product market fit reading when you split it. Calculate the repeat share once for customers who came from the first circle (the founder's followers, friends and launch list) and once for customers who found the brand through search, ads, marketplaces or press. If the second group comes back at a rate close to the first, the brand has a market beyond its audience. If only the first circle returns, the brand still has problem-solution fit, and the next launches should be designed to change that.

Reading those two shares from your own admin, with the cohort curve and the payback period next to them, is part of what The Conversion Audit maps in five business days.

Work through this with your own numbers

You run a creative brand that sells through [drops / a catalog / a membership / seasons], with [orders in the last 12 months] orders from [unique customers] customers, a [average order value] average order and a [gross margin] gross margin. Calculate the repeat share as orders minus unique customers, divided by orders. Then split it between [customers from the founder's followers, friends and launch list] and [customers from search, ads, marketplaces or press]. For the first-time buyers of your last three launches, estimate the share who ordered again within 90 days. Calculate the payback period as the cost to acquire a customer, [CAC], divided by the monthly gross margin per customer. Tell me whether the evidence points to problem-solution fit or product-market fit, and design one launch experiment, with a threshold written in advance, to move the weakest of the four numbers.

FAQ

How do you measure product market fit?

Count the customers who come back without being chased. In a store, four numbers do it: the repeat share of orders, the 90-day repurchase rate of first-time buyers, repeat purchase by launch cohort and the payback period of a new customer. Sean Ellis's survey adds a qualitative check.

What is the difference between problem-solution fit and product market fit?

Problem-solution fit is evidence that customers care about the problem and that the value proposition addresses it; for a creative brand, the first launches selling through to the founder's circle. Product-market fit is evidence that the offer creates value in the market: people outside that circle buying and coming back on their own.

What is the Sean Ellis test?

A one-question survey: how would you feel if you could no longer use the product? Very disappointed, somewhat disappointed or not disappointed. Ellis found that companies with strong traction almost always passed 40 % of 'very disappointed' answers.

Can a brand with a loyal following still lack product market fit?

Yes. If the repeat buyers are almost all from the founder's followers, friends and launch list, the brand has an audience and problem-solution fit. Product-market fit shows when customers who found the brand through search, ads or marketplaces come back at a similar rate.

Reading your repeat share, your launch cohorts and your payback period from your own admin is what The Conversion Audit maps. Five business days, $500, and the map stays with you whether or not you hire anyone next.