Finding Repeat Buyers Among First-Time Customers

Previously we covered timing emails to each customer’s own buying pattern. That needs two or three orders per customer. This one covers the customers who have only bought once.

For many stores they’re the largest group on the list, and the one thing they bought is the only information you have about them.

A first order is all you know about a new customer

Until a second order, there’s no pattern to work from. What you do have is the first purchase itself, and some first purchases lead to a second order far more often than others. Those are your gateway products.

A gateway product isn’t the same as a best seller. Plenty of best sellers are bought once and never followed by anything.

Find the products that lead to second orders

Take the customers who bought for the first time in the last year, group them by what was in that first order, and check what share ordered again within 90 days.

These figures are illustrative. In this example the true crime paperback is a gateway product, nearly a third of those customers order again, and quickly. The cookbook isn’t. The gift voucher rarely brings the buyer back, but the person redeeming it is a new customer who hasn’t yet seen why your store is good, so their first experience is where the extra effort goes.

The table also gives you timing. The median is taken from the customers who did order again, so it tells you when a second order usually happens for each group: around day 38 for true crime, around day 71 for art history.

Send the email when that group usually buys again

Send the email that follows a first order at around that median gap, to customers who haven’t ordered again yet. For true crime that’s around day 38, for art history around day 71. A single schedule for every new customer reaches one of these groups too early and the other too late.

Let the first purchase shape the recommendation

Content follows the same rule as before: what they bought decides what they see. A customer who started with true crime gets more true crime. For gateway products that’s enough.

For products that rarely lead to a second order, test pointing those customers to a gateway product instead of more of the same. A cookbook buyer may not want another cookbook, but could take a true crime recommendation.

Know where the data misleads you

Discounted first orders. A product can look like a gateway only because it’s always on sale. Check repeat rates for discounted and full price first orders separately.

Gifts and vouchers. One order creates two customers. The buyer’s order says little about what they’ll buy next, and the recipient is a new customer whose first order should be tracked on its own. Give the recipient’s first experience extra attention, since the buyer is trusting you with it. Orders shipped to a different address are the easiest gifts to spot.

Bulk and institutional buyers. Multiple copies of one title, ordered repeatedly, is a different pattern. Take these out before calculating anything.

Thin data. Products with only a handful of first time buyers won’t give a reliable rate. Group them by category.

Turn gateway products into an automated campaign

Start with one question: of the customers who bought for the first time in the last year, what share ordered again within 90 days, by first purchase? That gives you your gateway products. Then set one email for each, timed at that product’s typical gap.

90 days is a reasonable window to start with. A consumables business might use 30 to 45 days, a business selling considered purchases might use 180.

If you’d like this tracked automatically, that’s a conversation we’d like to have with you.