Average Revenue Per User (ARPU)

ARPU is monthly recurring revenue divided by the count of active paying users in that month. Not accounts. Not customers. Users — the individual seats, licenses, or logins that generate revenue. A company with $2 million in MRR and 8,000 paying seats runs a $250 ARPU. That number goes into every board deck for a seat-based SaaS company, and almost nobody in the room agrees on what "active user" means.

How ARPU Is Calculated

The formula is simple. The denominator is where companies quietly disagree.

$$ARPU = \frac{Total\ Recurring\ Revenue}{Number\ of\ Active\ Paying\ Users}$$

"Active" can mean logged in once in the trailing 30 days, or it can mean licensed regardless of usage. Those two definitions produce different ARPUs from the same revenue base, and finance teams pick whichever one flatters the trend line that quarter.

ARPU vs. ARPA — A Worked Example

Average Revenue Per Account measures the same revenue against a different denominator: accounts instead of users. The gap between the two numbers tells you how seat-dense your customer base is.

Same $2 million MRR, but now split across 500 accounts averaging 16 seats each:

An ARPA sixteen times the ARPU tells you the company sells into organizations, not individuals — a very different go-to-market motion than a $250-ARPU, self-serve product where accounts average 1.2 seats.

When Sales Teams Use ARPU

Product-led and seat-based SaaS companies live on this number. Pricing teams use ARPU trend to decide whether a tier restructure grew or cannibalized revenue. RevOps models expansion capacity off ARPU by cohort — a cohort with rising ARPU is buying more seats or upgrading tier, which is the cheapest form of expansion revenue a company can generate. Investors comparing two SaaS businesses of similar MRR use ARPU to infer whether one is selling enterprise-wide deployments and the other is selling individual licenses, which changes the entire valuation conversation about total addressable market.

Common ARPU Gaming Patterns

ARPU moves easily, which is exactly the problem. Widen the "active user" window from 30 days to 90 and dormant accounts count again, diluting a declining ARPU back toward flat. Bundle a free add-on seat into every paid contract and the user count balloons while revenue holds steady — ARPU drops, and the company reports it as a pricing win ("more value per dollar") instead of what it actually is, which is a discount.

The reverse trick works too. Strip out trial users, freemium accounts, and inactive seats from the denominator right before a board meeting, and ARPU jumps without a single dollar of new revenue booked. Multi-year contracts distort the monthly figure further — a three-year deal paid annually shows a revenue spike in year one that inflates ARPU for every user on the platform that month, then looks like churn when it normalizes in year two.

None of this shows up in the topline number. ARPU tells you monetization density per user, nothing about net revenue retention, nothing about concentration risk if 80% of revenue sits in 5% of accounts, and nothing about whether the users behind that revenue are actually using the product or just haven't been offboarded yet.