Time between purchases, sometimes called purchase cadence or interpurchase time, is the typical gap between one order and the next from the same customer. It is purchase frequency expressed as a duration, and it is the more actionable of the two.
For each customer, measure the days between consecutive orders, then take the median across customers.
A customer ordering on January 3, February 20, and April 2 has gaps of 48 and 41 days, giving a personal cadence of about 45 days.
Use the median rather than the mean. A small group of customers with very long gaps drags the average upward and produces a cadence that describes nobody.
The metric only exists for customers with at least two purchases, which means it silently excludes one-time buyers. That exclusion is intended, since it describes the people in a relationship, but it has to be stated, because a brand with poor repeat behavior can show a healthy cadence among the few who return.
Frequency tells you how often people buy. Cadence tells you when to act. A brand that knows its median gap is 45 days knows that a member at day 60 is behaving unusually, and can reach them while the relationship is still live rather than after it has lapsed. A member who goes quiet is harder to win back than someone who never joined, so acting early matters.
It is the basis of replenishment timing, win-back triggers, and at-risk definitions. Most programs that define an "at-risk member" are really defining a multiple of this number.
The target is a gap that matches how fast the product is actually consumed, with less variance around it.
Falling variance usually matters more than a falling median. Customers converging on a predictable cadence is what makes forecasting, inventory, and timed campaigns work.
Watch the distribution rather than the single number. And when you want to know what the program changed, test it against a small holdout group of members.
A supplements brand finds a median cadence of 38 days against a 30-day supply. The seven-day gap is the window where members run out and may buy whatever is nearest. Timing the reminder at day 26 rather than day 40 is the whole intervention.
A common mistake is using the mean. One customer who returns after three years can move the average for an entire cohort, and the resulting number will be wrong for every actual member.
Where TrueLoyal® fits: TrueLoyal flags each member's churn risk, and campaigns can be sent to a defined member segment, so a brand can reach members whose gap is stretching with a reminder or a winback offer before they lapse. Sending it to most of that group and holding a few back shows whether the timing actually brought people back.
Measure the days between each customer's consecutive orders, then take the median across all customers with two or more purchases. The median is the right choice because a few very long gaps distort the mean.
They are the same information in different units. Frequency counts purchases per period; time between purchases expresses the gap as days. Cadence is more useful operationally because it tells you when a member is overdue.
Most at-risk definitions are a multiple of the median cadence, for example a member whose last purchase was more than twice the normal gap ago. That gives a trigger based on the brand's own data rather than an arbitrary 30 or 90 day rule.
No. The metric needs at least two purchases to measure a gap, so one-time buyers are excluded by definition. That should be stated alongside the figure, because a brand with weak repeat behavior can still show a healthy cadence among the minority who return.
Purchase frequency is how many times an average customer buys in a defined period. With average order value and customer lifespan, it is one of the three numbers that produce customer lifetime value.
Time to second purchase is the number of days between a customer's first order and their second. It is measured on its own because the gap after a first purchase behaves differently from every gap after that.