## What is a Purchase Rate?

Aliases: *transaction rate*

Purchase rate is the proportion of purchases versus a base metric such as users, sessions, email subscribers, etc. with a generic formula being PR = P/N where P is the number of purchases and N is the number of events during which a conversion could have occurred. It can be expressed as a proportion or percentage, for example a user-based purchase rate can be calculated from 5 purchases out of 100 users as 5/100 = 1/20 = 0.05 = 5%. It is also often called transaction rate, or more specifically transaction rate per user, transaction rate per session and so on.

Purchase rate is often used as a Primary KPI since being a binomial metric makes it easy to work with due to its analytically known standard deviation. It should be noted that if a significant proportion of users make more than one purchase during a test this can violate the assumption of independence of observations so the parameter is no longer an independent variable. Significant violation leads to a compromised test and call for using more advanced methods that can account for the dependency.

If the revenue per purchase is equal or about equal across transactions the purchase rate can be used as the metric closest to the business bottom-line. If revenue per transaction varies significantly the conversion rate becomes a proxy metric and one should consider measuring average revenue per user (ARPU) instead despite the increased practical difficulty.

Like this glossary entry? For an in-depth and comprehensive reading on A/B testing stats, check out the book "Statistical Methods in Online A/B Testing" by the author of this glossary, Georgi Georgiev.