Every growth plan for an online store rests on an assumption about returning customers: how much they bring, and when. We took 36,000 orders, checked for each one whether the customer had bought from the store before, and followed the ones who came back. The answer changes where the money goes.
A returning customer is a customer who orders from the store again after their first order. The definition is simple, and the number behind it is almost never measured properly: most reports count "returning" only within the period they look at, so anyone who bought before the period started is counted as new.
We did it differently. We took 36,000 orders from online stores, January to September 2026, and for each order checked whether that customer had ever bought from the store before, in 2025 or earlier. A store whose history was too short to answer that question was left out. Revenue is the total of paid orders, after cancellations, the same way every Preneuriat dashboard counts it, and every figure is the median across stores, so no single large store tilts the picture.
In the median store, 22% of orders and 18% of revenue came from returning customers. In other words, 82% of the money came from people buying from the store for the first time.
The range is wide, from 8% to 31%, and what separates the ends is mostly the kind of product. A product that runs out and gets bought again, like skincare or food, builds a base of returning customers. A product that solves a need once, like furniture or a gift, does not. So there is no single "good" returning-customer rate. There is a rate that fits the product, and what matters is knowing yours and building around it.
The first conclusion is uncomfortable for anyone selling "retention" as a growth plan: in most online stores, growth is bought with new customers. In stores with a full year of history before 2026, customers acquired in earlier years brought only 10% of this year's revenue. 90% came from customers acquired this year. The base compounds, but slowly.
This is the finding that changes the work plan most. We followed every customer for 90 days after their first order, after their second and after their third, and asked how many ordered again inside that window.
After the first order, 7% of customers came back within 90 days. After the second, 15%, 2.2 times as many. After the third, 27%. Each additional order roughly doubles the odds of the next one.
A customer who has bought twice is not a "returning customer". They are a different customer altogether.
What this means for whoever builds the marketing: most of the money in loyalty programmes goes to customers who were coming back anyway. The place where one dollar does the most is the path from the first order to the second. That is where the odds are lowest, and where every small improvement changes the whole chain after it. We also looked at what brings a customer back from the product side, in our article on hero products: the hero product brings the first order, not the second.
Of the customers who bought for the first time between January and March, 11% came back within six months. Among those who did, the second order arrived after 48 days at the median. But the median hides the shape.
38% of second orders arrived in the first month, and 68% within three months. There is a clear wave in the first week, 15%, partly people adding something they forgot, partly customers whose parcel arrived and who wanted more. After 90 days the odds fall, but they do not vanish: almost a third of second orders arrived between three and six months after the first.
That says two things about post-purchase communication. First: the first month is the window, and one "thanks for your order" email does not fill it. Second: whoever stops talking to a customer after two months gives up a third of the second orders.
We expected a customer who already knows the brand to buy a bigger basket and pay full price. In the data, that is not what happened.
| Measure | New customer | Returning customer |
|---|---|---|
| Share of orders | 78% | 22% |
| Share of revenue | 82% | 18% |
| Order valueagainst a new customer's order | baseline | 4% less |
| Items per orderagainst a new customer's order | baseline | 8% more |
| Orders with a discountany discount on the order | 51% | 54% |
A returning customer's order was 4% smaller than a new customer's at the median, even though it held slightly more items: they buy more things, cheaper ones. And 54% of returning orders came with a discount, against 51% of first orders. The customer who knows the brand also knows its promotions.
This is not an argument against returning customers. It is an argument against the assumption that they are "cheap" for the store. A repeat order saves the cost of acquisition, and that is its real profit. The basket and the price it does not improve on its own.
In the online stores we measured, 18% of the median store's revenue came from customers who had bought there before, and the range across stores ran from 8% to 31%. 82% came from new customers. The number depends mostly on the kind of product: a product that runs out and gets bought again builds a larger base of returning customers.
A returning customer is a customer who orders from the store again after their first order. To measure it properly you have to check the store's whole history, not only the period you are looking at, otherwise anyone who bought before the period started is counted as a new customer.
In the median store, 7% of customers ordered again within 90 days of their first order, and 11% within six months. After the second order the odds of another order within 90 days rose to 15%, and after the third to 27%. A good rate is one that fits your product; what matters more is that the rate climbs after each order.
Among customers who came back within six months, the second order arrived after 48 days at the median. 38% of second orders arrived in the first month, 68% within three months, and the rest between three and six months.
Not in our data. A returning customer's order was 4% smaller than a new customer's at the median, with slightly more items in it, and 54% of returning orders carried a discount, against 51% of first orders. A bigger basket has to be built into the offer.
ROAS is the number every ad platform reports and every ecommerce brand argues about. We set what Meta and Google reported against what online stores actually sold over nine months of 2026. Together, the two platforms took credit for 106% of everything the stores sold.
Daniel Hirsh30 September 202612 min readEveryone watches the CTR to know when an ad is done. We took 271 image ads in Israeli online stores apart, week by week from their first day, to see what really wears out, when, and what that means for when to change the image.
Michael Mervinetsky21 October 202610 min readThe night sale is a very common promotion in ecommerce, and almost nobody measures what it does beyond the evening itself. We took 9 night sales apart, hour by hour, to see when the money actually comes in and what happens to the days around it.
Daniel Hirsh18 October 20267 min readEvery Preneuriat client gets a live dashboard: store revenue, all media spend, real ROAS, updated all day. The first call goes both ways.
Check the fit
