Average Revenue per User

"Revenue doubled" hides two completely different events behind one number. Either twice as many people are paying, or the same people are each paying twice as much. The first usually comes from spending more on ads; the second comes from changing the product or the pricing. What the company should do next runs in opposite directions depending on which one happened, and a table of totals shows both as the same taller bar.

Due diligence meetings used to stall in exactly that spot. Grace would say revenue grew 40 percent for the quarter, and someone across the table would ask how much the user count grew. When the answer was 55 percent, the room went quiet. Revenue was up while the amount each person paid had gone down, which meant growth bought with money was covering for growth the product hadn't earned.

Average revenue per user (ARPU) exists to pull those two events apart. Divide the total by the head count and growth splits into two lines: one says how many people arrived, the other says whether those people pay more than the ones before them. Neither line is the whole story on its own, and a company can look healthy on one while quietly failing on the other.

Divide recurring revenue for a period by the number of users in that same period. Most teams divide monthly recurring revenue (MRR) by the paying users that month and read it monthly.

MRR is the sum of what is contracted to come in again next month. One-time setup fees, data migration charges, and consulting work drop out, because they won't repeat. Clean that up first or the figure stops describing a level the company can hold: one big implementation fee lands, the number spikes, and the plan built on that spike misses immediately.

The calculation has three steps.

1. Fix the numerator. Use MRR for the month. Pagedeck's was $40,000 last month.
2. Fix the denominator. Count users, but agree first on whether people on the free plan are in or out. Pagedeck has 2,000 paying users and 20,000 including the free plan.
3. Divide. On paying users that's $20. Counting everyone, it's $2.

Same company, same month, a tenfold gap. The paying-user version answers "what is one paying customer worth," so it drives pricing decisions. The all-users version answers "what does this whole audience convert to," which is the scorecard for freemium work — freemium means opening the basic product for free and charging only for the advanced parts. That is why the first question about this number is always which denominator produced it.

Keep the time unit straight too. Multiply a monthly figure by twelve and you get an annual one, and an annual figure parked next to a monthly one with no label makes a twelvefold difference look like growth. One unit per table.

There is no benchmark. Products people buy individually and products companies sign contracts for don't share an order of magnitude. Consumer collaboration tools commonly land between $10 and $25 a month, while a product built on large enterprise deals can pass $75 per user. Compare against your own last twelve months rather than the company next door. A figure that held flat for a year while the user base tripled is still a result worth reporting; the same flat figure with a shrinking user base is not.

Read direction and cause together. The number rises for three reasons: prices went up, existing customers moved to a higher plan, or cheap-plan users left and lifted the average of whoever remained. The first two are good news, the third is bad news, and all three draw the same upward line. So a report that ARPU rose comes with the user-count chart attached. If the count is falling while the average climbs, it's the third one.

Remember it's an average. If 20 of 2,000 users pay $1,500 a month and the rest pay $8, the average comes out close to $23 while almost nobody actually sits there. Lose a handful of those 20 and the average collapses within days. Looking at the median alongside the mean, or splitting the figure by plan, exposes that concentration. Teams that only ever chart the average tend to find out about it the month a large account cancels.

"Revenue is up but ARPU slid from $22 to $20."

Growth came from adding people. Stop the ad spend and the growth stops with it.

"Does that number include free users?"

Checking the denominator. The answer moves the figure by a factor of ten.

"The pricing revamp targets a 15 percent lift in ARPU."

A plan to grow revenue by raising what each person pays instead of adding people.

At Pagedeck this is the second slide in the monthly metrics review. Ravi reports the month's new signups and Mia reads ARPU next to it. If signups rise and the average holds, the same kind of user is arriving in greater numbers. If signups rise while the average drops, the new arrivals are crowding into the cheapest plan. Grace uses those two lines to decide where next month's budget goes.

A low number doesn't mean raise prices tomorrow. Moving customers to a higher tier or getting them to add seats works too, and the people who talk to customers know which of the three is available.

The most common mistake is reading any increase as good news. An upward line looks like an upward line. But this number also rises when cheap-plan users leave in bulk, because the ones who stayed were on pricier plans. Report that as an achievement and the churn stays hidden for months, until total revenue starts falling and someone finally goes looking for the cause.

Reading it as profit is just as common. Someone paying $25 a month is not $25 the company keeps. Hosting, support salaries, and payment fees come out of it, and the customer acquisition cost (CAC) spent to win that person still has to be earned back. CAC is ad spend plus sales salaries divided by the number of new customers. At $25 of ARPU against $150 of CAC, it takes six months to break even, and if people leave before that, every additional user costs the company money.