Average Revenue per Account

In a subscription business, knowing that this month brought in $38,000 tells you almost nothing about the company. A hundred accounts paying $380 each and ten accounts paying $3,800 each produce the same figure and describe entirely different businesses. Lose one customer in the first and you lose 1 percent; lose one in the second and 10 percent is gone. How you sell, how many support people you staff, and what the price list looks like all differ.

Before anyone drew that distinction, investor conversations went in circles. Put two growth rates side by side and the companies look identical, when in reality one of them has its fate tied to two renewal conversations and the other doesn't. That's why "how many customers is that?" always came next.

In a subscription business selling to companies, the unit of a contract is an organization, not a person. One contract might cover fifty seats or three. Divide revenue by headcount and contracts with many seats blur together with contracts with few, and the resulting number stops pointing at anything. Average revenue per account, usually shortened to ARPA, puts the denominator back on the entity that actually signs. It's the number you reach for when setting a price list or deciding what size of customer to chase.

Divide monthly recurring revenue (MRR) by the number of accounts that month. MRR is the sum of what is contracted to arrive again next month, and one-time money such as implementation or consulting fees stays out of it.

1. Find recurring revenue for the period. Annual contracts get divided across twelve months, so a $9,000 annual deal counts as $750 this month.
2. Count the accounts. Count only paying contracts and leave free accounts out, or the number swings for no reason — put free accounts in the denominator and the figure falls even in a month when paying customers grow.
3. Divide. With $38,000 in MRR across 100 accounts, ARPA is $380.
4. For an annual view, divide annual recurring revenue (ARR) by the same account count. The business above has $456,000 in ARR, so the annual figure is $4,560. Mixing monthly and annual in one meeting makes the scale look twelve times off, so name which one you're quoting first.

Writing down the counting rule is the genuinely hard part. If one company signs two contracts through two departments, is that one account or two? Does a customer who started on the 20th belong in this month's denominator? Without a decision, finance quotes one number and sales quotes another, and the meeting goes to reconciling them. Write one line — paid accounts as of month end — and count the same way every month after.

There is no benchmark for the absolute value. A business selling to individuals and small teams runs around $25 a month; a business signing large enterprises runs around $15,000. Neither is better. The only question is whether the figure matches the model you chose.

Direction is where the meaning is. If the same business moves from $310 to $425, existing customers added seats, moved to a higher tier, or the price went up — and which of the three it was only shows once you break revenue change into its parts.

A drop isn't automatically bad either. Launch a cheap small-team plan and account count climbs fast while the average sinks. What to check then is whether total recurring revenue rose alongside it. If it did, the market got wider. If revenue held flat while the average fell, existing customers moved down to the cheaper plan.

The figure to read next to it is what one account costs to win. When customer acquisition cost (CAC) is $4,600 and ARPA is $380, even the crude version of the math needs twelve months before the money spent comes back. That's the situation where raising this number earns its keep: a payback period that's too long leaves you two moves, lowering acquisition cost or raising revenue per account.

"ARPA has been stuck at $310 for two quarters."

Account count is growing while what each one pays stays flat.

"Look at that per account, not per user. Per user buries the customers with a lot of seats."

Using headcount as the denominator hides the effect of the large contracts.

"Put free teams in the denominator and the number halves."

Two people are applying different counting rules.

Grace checks this figure at every monthly metrics review at Pagedeck. Last quarter recurring revenue grew 20 percent while ARPA slid from $370 to $340. Her question was which customers had arrived in the meantime.

Sofia pulled the list: of the 60 new accounts, 47 were teams of five or fewer, all of them arriving through the small-team plan launched in March. Not one existing customer had reduced spend, revenue was up, and only the average had moved. Rather than trying to push the average back, the team decided to track seat growth within those small accounts separately. A team that starts at five and reaches twelve within six months pulls the average up on its own later.

The most common reading is that a higher number means a better business. Receiving more from each account looks that way. But the fastest route to a higher number is to stop accepting small customers, which shrinks the denominator without adding revenue — the average rises while the total holds flat or falls. Chase only large contracts and the customer count gets thin enough that a single departure moves the whole company. What deserves to rise isn't this figure but what's left after the cost of winning the account.

Reading past the word average is the other frequent mistake. This is a sum divided by a count, so a handful of large customers lift the whole thing. If three of 100 accounts pay $15,000 a month and the other 97 pay $150, the average is about $600 while the most common customer is a $150 one. Redraw the price list around $600 and the band 97 accounts were living in disappears. Split the figure by customer size, or look at the account sitting in the middle, and the trap is visible.