Definition
ARPPU (Average Revenue Per Paying User) is a defined revenue numerator divided by the number of paying active players in the same period. It describes the mean among payers; it does not by itself establish channel quality, profitability or player value.
Formula
ARPPU = NGR ÷ PAU
NGR must follow the operator's documented revenue policy; bonus, tax, payment and supplier-cost treatment can differ (see GGR and NGR). A GGR-based and an NGR-based ARPPU are different metrics, so the numerator must be named. PAU — paying active users: distinct players meeting the documented paying event in the period.
When ARPU and ARPPU use the same non-negative numerator, period and activity rules, and payers are a subset of actives, ARPPU is at least as high as ARPU. If those conditions do not hold, the relationship cannot be inferred.
How operators use it
| Reporting view | Required context |
|---|---|
| Blended ARPPU | PAU count, currency, period and documented NGR policy |
| Segment ARPPU | Stable segment definition, cohort age, product and source |
| Distribution | Median, percentiles and concentration as well as the mean |
ARPPU can support channel and segment analysis when period, revenue basis, cohort age, product mix and payer definition are held constant. Read it with cost per first-time depositor, retention, distribution and contribution margin; a mean alone cannot distinguish durable value from a small number of concentrated payers.
Because the paying base can be heavily skewed, ARPPU should be read by segment as well as in aggregate. Blended ARPPU can rise because mid-value players churned, shrinking the denominator faster than revenue — a deterioration disguised as improvement. Reporting should therefore split ARPPU by stable loyalty tier, cohort age and product, and show medians and distribution shifts alongside the mean.
ARPPU can be one input to an LTV model, alongside payer probability, retention, costs and the chosen revenue definition. Changes in ARPPU, payer mix or retention should trigger model reconciliation against realised cohort contribution rather than an automatic LTV conclusion.
Common misinterpretations
A common error is publishing ARPPU without its revenue policy: GGR- and NGR-based versions are not interchangeable. Another is confusing ARPPU with average first deposit: one is periodic revenue per payer and the other is a transaction measure. A rising blended ARPPU is also not automatically good news; check whether revenue grew, the payer base shrank or concentration changed.
Related metrics: ARPU · Paying Active Users · LTV · GGR and NGR · Bonus Cost
Common questions
What is the difference between ARPU and ARPPU?
ARPU divides a defined revenue numerator by all active players; ARPPU divides the same numerator by paying active players. When both use the same non-negative numerator and paying users are a subset of actives, ARPPU is at least as high as ARPU. The comparison is invalid if revenue basis, period or player definitions differ.
Should ARPPU be calculated on GGR or NGR?
On NGR. Bonus cost is real money given back to players, and at meaningful bonus intensity a GGR-based ARPPU materially overstates player value — precisely in the segments and channels where bonusing is heaviest and the truth matters most.
What is a typical ARPPU for an online casino?
There is no defensible universal amount. Currency, market tier, product, payer definition, NGR policy, cohort age and high-value-player concentration can move ARPPU materially. Establish a like-for-like internal baseline and report the median, percentiles and payer count alongside the mean.