Repeat purchase rate and purchase frequency

Repeat purchase rate is the share of customers who buy again. Purchase frequency is how often they do it over a period. Together they’re the foundation a DTC business stands on, because first-order economics mean the brand only works if customers come back, and these two numbers are how you know whether they will.

The reason this is the foundational metric rather than lifetime value itself is timing. LTV is a backward-looking total that takes a year or more to mature, which makes it useless for decisions you need to make now. Repeat rate shows up early. Measure the share of a cohort that places a second order within 30, 60 and 90 days and you have a read on LTV months before LTV itself is knowable. That early read is what lets you set acquisition budgets and judge tests before the losses pile up.

Reading it as a cohort, not an average

A blended repeat rate across all customers is close to meaningless, because it mixes cohorts at different ages. A customer acquired last week hasn’t had time to reorder; one from last year has. The honest version groups customers by acquisition month and tracks each cohort’s repeat behaviour over time, the same cohort logic retention analysis uses in SaaS.

What you’re looking for in the cohort curves:

• The second-order rate. The gap between first and second purchase is the steepest drop-off and the most predictive moment. A customer who orders twice is dramatically more likely to order again.
• Time to second order. This tells you how long your acquisition payback actually takes, and whether your category has a natural reorder rhythm.
• Whether cohorts are improving. A brand getting better at retention shows it in newer cohorts reordering faster than older ones did.

Things people get wrong

• Confusing repeat rate with retention rate. Repeat rate counts whether someone bought again. It says nothing about how many one-time buyers are sitting in the base dragging the average down.
• Optimising acquisition for cost per first order. The cheapest first orders often have the worst repeat rates, because discount-led and bargain-hunting traffic converts once and leaves. Cheap acquisition that doesn’t repeat is expensive.
• Reading it without segmenting by product or source. A consumable and a one-off durable have completely different natural frequencies, and lumping them gives a number that describes no real customer. Use segmentation to see the customers underneath the average.