ARPAAverage Revenue Per Account
Average monthly revenue per paying account, not per seat.
Put this metric to work: explore FlowMRR's recurring revenue analytics to track revenue movements, retention and forecasts from your Stripe account, or see how FlowMRR compares as a Baremetrics alternative for Stripe.
ARPA measures monthly revenue per paying customer account. FlowMRR does not compute a separate ARPA: its ARPU divides MRR by paying people (one per email, Stripe customer ID as fallback), which equals ARPA when each paying person is one account. If ARPU uses seats as its denominator, the two diverge: one account can hold dozens of seats.
Formula
ARPA = MRR ÷ paying accountsOther definitions you will see
Some tools use ARPA and ARPU interchangeably for revenue per Stripe customer record. Others compute ARPA over every account, free ones included, or on an annual basis close to ACV. Per-seat ARPU, common in B2B, divides by seats instead of accounts and gives a much lower figure.
Illustrative comparisons
By motionThese unsourced ranges are illustrative scenarios, not measured industry benchmarks, percentiles or recommended targets. Their qualitative labels are editorial interpretations. Compare your own cohorts using the same segment, period, currency and measurement basis.
Why it matters
ARPA is the number that should trend up as you move up-market or layer in expansion. Rising ARPA with flat customer count means your existing base is paying you more — the cheapest growth there is. Confusing ARPA with per-seat ARPU is how teams misprice multi-seat plans.
Common mistakes
- Using ARPU and ARPA interchangeably when your plans are seat-based.
- Computing ARPA on total revenue including one-off fees.
- Not segmenting ARPA by plan or cohort, which hides where expansion actually happens.
Calculating it from Stripe data
Stripe has customers, not accounts, so the first step is deciding what an account is. A Stripe customer record is the simplest unit but often the wrong one: one company can pay through several records (one per workspace, team or billing contact), and one person can end up with a second record after re-subscribing. Group records that belong to the same account, by a company or account ID you store in customer metadata if you have one, otherwise by email. FlowMRR counts paying customers once per person (email first, Stripe customer ID as fallback), so its ARPU is revenue per paying person; for a company that pays through several people's emails, the grouping has to come from your own account ID.
The numerator is MRR on the usual basis: monthly-normalised recurring prices, forever coupons deducted, trialing, incomplete, incomplete_expired and paused subscriptions excluded, unpaid ones out from the end of their last paid period. The denominator is accounts with MRR above zero in the same month. An account on a free plan, fully discounted or with collection paused pays nothing and belongs in neither side; an account that churned during the month is out of both.
If your prices are per seat, the subscription item's quantity gives you seats. MRR ÷ accounts gives ARPA, MRR ÷ total seats on seat-based prices gives revenue per seat, and seats ÷ accounts gives the average account size. Keeping all three apart shows whether a change came from price per seat or from accounts adding people.
Reading the result
MRR is accounts × ARPA, so every change in MRR splits into a change in the number of paying accounts and a change in what each pays. ARPA rises when accounts expand or when the mix shifts toward larger accounts, including when small ones leave. Check the account count and the ARPA of newly acquired accounts before reading a rise as pricing power: if new accounts start lower than the base, the average will drift down as they arrive even if no one downgrades.
Because a few large accounts can pull the mean far up, look at the distribution too: the median account, ARPA by plan and by signup cohort, and the share of MRR held by the largest accounts. ARPA by acquisition channel is also the figure to set against that channel's CAC.
Frequently asked questions
What is a good ARPA for SaaS?
There is no target that fits every business. ARPA follows your sales motion and customer size: a self-serve product for individuals and an enterprise product sold by a sales team can both be healthy at very different levels. What matters is that ARPA supports your acquisition cost, which is why it is read next to CAC and payback.
How do I count accounts for ARPA from Stripe?
Count distinct paying accounts in the month, not subscriptions and not raw customer records. Group Stripe customers that belong to the same company or person, by an account ID in metadata or by email, and leave out customers whose subscriptions bill nothing.
What is the difference between ARPA and ACV?
ARPA is monthly recurring revenue per paying account across the whole base. ACV, annual contract value, is usually the yearly value of a contract, often a newly signed one, and may include one-off fees depending on who reports it. ARPA × 12 equals annual revenue per account only on the same recurring basis.