Average Revenue per User

Average revenue per user (ARPU) is total revenue for a period divided by the average number of users in that period. It is a monetisation gauge, not a profit measure. It says what the base pays, not what serving it costs.

The formula is simple. The argument is about the denominator. Count every registered account and ARPU looks thin; count only paying accounts and the same revenue produces a much bigger number, usually reported as ARPPU.

ARPU is also a non-GAAP measure. No accounting standard defines it and no regulator polices it — so two companies can publish the same label and mean different things. That’s why the definition matters more than the figure.

Key takeaways

• ARPU is period revenue divided by the average user count for that period.
• Say which denominator you used: all users, or paying users only (ARPPU).
• Normalise the period, because monthly and annual ARPU are not interchangeable.
• ARPU tracks your own base well and compares across companies badly.

How it works

ARPU divides the revenue you booked in a period by the average number of users over that same period. Both halves need a stated definition: which revenue lines are in, and which accounts count as a user.

Start with the denominator. That’s where the arguments live. A freemium product with millions of free accounts and a thin paying minority reports a tiny ARPU. Split the same revenue across paying users only and you get ARPPU, a far bigger number.

Quoting ARPU for a freemium product without naming the denominator is close to meaningless. The number can move by an order of magnitude on definition alone — before a single customer changes behaviour.

The numerator needs the same discipline. Subscriptions only, or subscriptions plus advertising, hardware and add-ons? Widening the revenue definition lifts ARPU without anyone buying more, which is why the definition belongs next to the figure.

Then the user count itself. Use a mid-period average, not a period-end snapshot. When the base is growing fast, the closing count sits well above the average, so it drags reported ARPU down for no real reason.

Period normalisation matters just as much. Monthly ARPU and annual ARPU differ by roughly a factor of twelve, so never compare one against the other without converting first. State the window every time you publish.

Here’s the sequence in practice:

1. Pick the period and state it: monthly, quarterly, or annual.
2. Total the revenue you’re counting, and say which lines are in and which are out.
3. Average the user count across the period, ideally from monthly or daily snapshots.
4. Divide, then publish the definition next to the number so nobody has to guess.

Pair ARPU with churn and you get a rough lifetime value: divide ARPU by the monthly churn rate and you have the revenue an average user delivers before leaving. It’s an estimate — not a bank statement.

Segmented ARPU beats blended ARPU almost every time. Split by plan tier, region or cohort and you can see which slice is actually improving, instead of watching one number hide two opposite trends.

So ARPU is strong at one job and weak at another. It tracks how well you monetise a base you already have. It’s a poor referee between companies that count users differently.

ARPU sits outside GAAP, so the definition a listed company uses lives in its own filings. You can read those in the SEC’s EDGAR filing database, where US-listed firms file annual and quarterly reports.

Examples

ARPU shows its worth when you compare a business against its own past under a stable definition. Here are four situations worth walking through — three where the metric earns its place, and one where it misleads.

A freemium app. Most accounts pay nothing, so blended ARPU stays low while ARPPU looks healthy. Watch the free-to-paid conversion rate alongside both figures, or you can’t tell monetisation gains from signup growth.

An online retailer. The US Census Bureau put US retail e-commerce sales at $329.5 billion in the second quarter of 2026, up 12.4 percent year on year, and 17.1 percent of total retail on an adjusted basis.

A retailer in that market can compute ARPU per active shopper, but revenue per order and orders per shopper drive it. Sell more per visit through upselling and cross-selling and ARPU rises without a single new user.

A support-heavy service. ARPU only means something next to cost to serve. The US Bureau of Labor Statistics put the median hourly wage for customer service representatives at $20.59 in May 2024.

The same handbook projects employment in those roles to decline 5 percent from 2024 to 2034, with about 341,700 openings a year. Rising ARPU alongside falling support cost per user is the pattern worth chasing.

A cross-company comparison. This is where ARPU breaks. One firm counts monthly active users, another counts registered accounts, a third counts paying seats. Ranking them on ARPU compares three different denominators — and tells you almost nothing.

Across all four, the pattern holds: ARPU is a strong internal tracker and a weak external scoreboard. Keep the definition fixed, and the trend line will tell you something real.

Related terms

ARPU sits in a family of unit metrics that only make sense together. Each one measures a different slice of the same customer relationship, so read them as a set rather than picking a single favourite.

• Customer Lifetime Value: total revenue expected from one customer across the whole relationship.
• Average Order Value: mean revenue per transaction rather than per user per period.
• Customer Acquisition Cost: spend needed to win one new customer, the counterweight to ARPU.
• Customer Retention: share of users who stay, the input that turns ARPU into lifetime value.
• SaaS: software sold by subscription, the model where ARPU reporting is most common.

FAQ

What is average revenue per user?

It’s total revenue for a period divided by the average number of users in that period. The figure tells you how well you monetise the base you already have, not how profitable that base is.

What’s the difference between ARPU and ARPPU?

ARPU divides revenue by all users, while ARPPU divides the same revenue by paying users only. For a freemium product the two can differ by an order of magnitude. Always label which one you’re quoting.

Should I report monthly or annual ARPU?

Either works, as long as you state the window and stick to it. Monthly and annual figures aren’t interchangeable, and switching between them without saying so is how comparisons go wrong.

Why use an average user count instead of the period-end number?

Because a base that grows fast during the period makes the closing count much larger than the true average. That inflates the denominator and pushes reported ARPU below what your users actually paid you.

Is ARPU a GAAP measure?

No, ARPU is a non-GAAP measure with no standard definition, which is why any company reporting it should publish that definition beside the number.

Browse the Outsource Accelerator directory to find partners who can grow your user base without inflating what it costs to serve them.