An online store in Peru closes the month happy: cost per acquisition dropped, sales went up, the ads dashboard looks better than ever. Nobody in the meeting asks whether those same people are going to buy again. The metric that carries the most weight in the decision to raise ad spend is, almost always, the one that says the least about whether the business actually works.
The first purchase is bought by the algorithm, the second one is earned
Peru’s ecommerce market is growing fast, but it still starts from a small base: a large share of today’s online buyers are trying it out for the first or second time, not repeating a years-old habit. That means a well-built campaign gets first purchases without much effort, because new demand keeps entering the channel all the time. The second purchase is a different story: from there, the algorithm has no say, and it is the product, the shipping experience and the post-purchase communication that decide whether the buyer had a real reason to come back.
The usual mistake is measuring the business only by conversion rate and cost per acquisition, two numbers calculated on the first purchase. A month with good conversion can be hiding a low repurchase rate, which means paying the full acquisition cost every time, including for the sale that should have come almost free, from someone who already trusts the brand.
What to look at besides conversion
- Measure repurchase by cohort, not as one aggregate. Group buyers by the month of their first purchase and check what share came back at 30, 60 and 90 days. One blended number hides the difference between a good cohort and a bad one.
- Check what reaches the customer between purchase one and two. If no message is triggered by that first purchase (confirmation, follow-up, a repurchase nudge), the second sale depends on the customer remembering the brand on their own.
- Compare the cost of acquiring a new customer against the cost of winning back one who already bought. The second is almost always cheaper, but only if a routine actually attempts it.
- Separate product from communication. Low repurchase can mean the product did not create a second need, or that nobody told the customer they could buy again. Those are two different problems with two different fixes.
It is the same standard we apply in growth for brands operating in Peru: before raising acquisition budget, confirm the business retains what it already won. And it connects directly to the lifecycle-stage retention framework: the second purchase is, literally, the first step in that framework, the one that separates a customer with zero purchases from one with a single purchase.
If your monthly report has cost per acquisition and conversion rate but no repurchase rate, you are probably measuring how well the algorithm buys, not how well your business sells.