Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Metric

Purchase Frequency

Average number of orders each customer placed in the period.
BusinessmetricUpdated

Formula

Orders in the period ÷ Unique customers in the period

Unit: ratio · higher is better

Definition

Purchase frequency counts how often customers bought, not how many customers there were. Over a chosen period, take every order placed and divide by the number of distinct customers behind those orders. A customer who ordered once counts once in the denominator; a customer who ordered six times also counts once, and contributes six to the numerator. The period has to be stated with the figure, because a quarterly frequency and an annual frequency describe the same customers at different scales.

Orders in the period ÷ Unique customers in the period(ratio)

  • Orders in the period are completed orders, usually after cancellations. Whether returned orders stay in is a choice to state; most teams keep them, because the customer still chose to buy.
  • Unique customers in the period are distinct buyers, identified by account, email or a matched customer ID. If guest checkouts cannot be joined to earlier orders, one person appears as several customers and the ratio falls for reasons that have nothing to do with behaviour.
  • The period is fixed in advance and should be long enough to cover the normal gap between orders in the category. A month is meaningless for a product people buy twice a year.

The denominator is customers who bought in the period, not every customer on the database. Dividing by the whole list mixes lapsed customers into the base and produces a much lower number that answers a different question.

How to interpret it

Higher is better, but the single average says less than the shape behind it. Three readings make it useful:

  • The distribution. Count how many customers placed one order, two, three and so on. Most retailers find a long tail of one-time buyers and a small core of regulars. The mean sits between them and describes neither group, so a plan aimed at "the average customer" can miss both.
  • By cohort. Customers acquired in the same month, followed for the same length of time, show whether newer customers come back as often as older ones did at the same age. A blended figure is propped up by long-standing loyal buyers.
  • By category. Consumables, gifts and durable goods have different natural cycles. Comparing frequency across categories, or against a business in a different one, compares replenishment cycles rather than marketing.

The jump that matters most is usually from one order to two. The repeat purchase and loyalty guide treats that second order as the moment a buyer becomes a customer, and this metric is how its effect shows up across the base.

Illustrative example — not a real result

A coffee roaster looks behind its average

Rookwood Coffee Roasters took 6,480 orders last year from 2,400 customers, a purchase frequency of 2.7. The team had planned a loyalty reward for its "typical" customer who orders about three times a year. Counting orders per customer told a different story: 1,560 customers ordered once, 600 ordered between two and five times (2,040 orders), and 240 subscribers ordered every month, placing 2,880 orders between them. The subscribers are 10% of customers and 44% of orders, so the average describes neither group. Rewarding the core would have paid people who already buy monthly. Rookwood moved its retention budget to a second-order offer for the 1,560 one-time buyers instead: if one in ten of them places a second order, that is 156 orders that would not otherwise have happened.

What moves it

  • Replenishment cycles. Products that run out on a predictable schedule set a natural rhythm; reminders timed to that rhythm raise frequency more than generic promotions.
  • Range breadth. A customer can only buy as often as there is something relevant to buy. Adjacent products and refills give returning customers a reason beyond the original item.
  • Reminders and lifecycle messages. Post-purchase and replenishment emails catch the moment a customer is ready to reorder; the lifecycle marketing guide sets out which stage each message serves.
  • Subscriptions and loyalty schemes. Subscriptions lift frequency mechanically; points schemes mostly reward behaviour that would have happened anyway unless they target the step from first to second order.
  • Experience of the first order. Late delivery, damage or a confusing return process lower the chance of any second order, which caps frequency before marketing can act.

What good looks like

There is no reliable benchmark, and the site does not cite one. A good frequency depends on what is sold, how quickly it is used up and how the business sells it, so a figure that is excellent for furniture would be a warning for pet food. The useful comparisons are internal: each cohort against earlier cohorts at the same age, the share of first-time buyers who place a second order, and the movement of the distribution over time. The economic test asks whether the frequency the business actually sees, multiplied by order value and margin, recovers what it costs to acquire a customer. The growth economics section works through that arithmetic.

Purchase frequency is one factor of LTV: order value times frequency times lifespan, with margin applied, gives the value of a customer. It pairs with AOV, because the two together make revenue per customer, and a promotion that raises one often lowers the other. Repeat Purchase Rate asks a narrower question, the share of customers who bought more than once, and is often the better tracking metric precisely because it ignores how many extra orders the heaviest buyers placed. Churn plays the same role for subscription models, and the churn and retention metrics guide shows how the retention family fits together. The habit-forming loops that raise frequency without paid media are covered in growth loops in practice, and the wider e-commerce retention section lists the programmes that act on it.

Common mistakes

  • Dividing by the whole customer list instead of the customers who bought in the period.
  • Quoting the mean alone, so a small loyal core hides a majority who never came back.
  • Choosing a period shorter than the buying cycle, which makes frequency look low and volatile.
  • Failing to join guest orders to known customers, which splits one buyer into several.
  • Comparing across categories with different natural cycles as though the gap were a marketing result.
  • Rewarding the heaviest buyers first when the larger gain sits in converting one-time buyers to a second order.

More terms to look up

Where this term is used

Related