Time Between Purchases Calculator

How the number is derived

Average time between purchases is the total days between first and last order divided by the number of intervals: (last order − first order) ÷ (orders − 1), averaged across customers with two or more orders.

The median is more useful than the mean, since a few very long gaps distort the average upward.

It defines the natural purchase cycle, which is what every retention timing decision should be set against.

An example

Repeat customers averaging 2.4 orders a year have a median interval of roughly 152 days, about five months.

That means a reminder at 90 days is early and one at 240 days is late. The window where a nudge changes the outcome is somewhere around 130 to 170 days.

It also defines lapsing: a customer at 300 days is genuinely overdue, where at 160 they are behaving normally. Flagging them at 90 would treat most of the healthy base as at risk.

For a business whose retention emails fire at a fixed 30 or 60 days, that mismatch means most of the sends reach people who were never going to buy yet.

Where it is unreliable

Averaging across products with different cycles produces a figure describing neither. A business selling both consumables and durables needs the calculation per category.

Customers who bought once are excluded by definition, which biases the figure toward the more engaged part of the base.

What this changes

Calculate it per product category and set replenishment timing from the specific cycle rather than from a site-wide average.

Then define lapsing as a multiple of the median, perhaps 1.5 or 2 times, so the flag fires when behaviour has genuinely changed rather than on a calendar convention.

Timing as the cheapest retention lever

A message arriving when the customer is ready to buy converts far better than the same message arriving early or late, and the difference is usually larger than anything the copy achieves.

That makes the purchase cycle one of the highest-value pieces of analysis available, and it requires only order history to compute.

Businesses that set reminder timing by convention rather than by measurement are sending the right message at the wrong moment, which produces poor results and is usually diagnosed as a content problem. Checking the interval first costs an hour and frequently fixes it.

Comparing the interval between a customer's first and second order against their later intervals shows whether the relationship accelerates, which is a useful signal about product fit.

Customers whose intervals shorten are becoming habitual; those whose intervals lengthen are drifting away before they formally lapse.

Setting automation timing from the median rather than the mean avoids the distortion caused by a small number of very long intervals.

Recalculating the interval annually catches shifts caused by product mix changes, since adding or removing a category moves the natural cycle.

Building the interval into a subscription or auto-reorder offer converts a timing problem into an operational one the customer no longer has to solve.

Comparing the interval for customers acquired through different channels sometimes reveals that the cheapest traffic also buys least often, which changes the true cost per order.