ARPA (Average Revenue Per Account)
Recurring revenue divided by number of accounts. The per-customer economics figure that decodes what is driving ARR growth.
What it measures
ARPA — average revenue per account — is recurring revenue divided by the number of accounts producing it, over a stated period. It is a headline efficiency and positioning number: it says roughly what a customer is worth per period, and its direction says whether you are moving up-market or down.
ARPA = recurring revenue for the period ÷ number of active accounts in that period. The definitional work is entirely in the denominator — whether it counts accounts, billing entities or logos changes the answer materially in any business where one customer holds several contracts.
Blended ARPA hides the story
How to make it useful
Segment before averaging. ARPA by plan tier or by segment tells you something; a single blended figure across a range of customer sizes mostly measures mix.
Report it alongside account count. ARPA rising while accounts fall is churn at the small end, not success at the large end, and the two require opposite responses.
Compute new-customer ARPA separately from base ARPA. New-customer ARPA is the leading indicator of where the business is heading; blended ARPA is a lagging average of where it has been.
Fix the denominator definition in writing. Accounts, logos and billing entities produce different numbers, and drift between them shows up as a trend nobody can explain.
Look at the distribution periodically. A mean pulled by a handful of very large accounts describes almost none of your customers.
Where it misleads
It is a mean, so it is sensitive to outliers. In most businesses the median account is materially smaller than ARPA suggests.
Rising ARPA through churn of small accounts is a deterioration that reads as an improvement.
Mixing one-off and recurring revenue in the numerator makes the figure jump with services delivery rather than with the subscription base.
ARPA cannot distinguish a price increase from a mix shift, and the two have different durability — one is defensible, the other is a change in who your customers are.
RELATED TERMS
Annual Recurring Revenue (ARR)
The annualised value of recurring subscription revenue at a point in time. A run rate, not a measure of revenue earned during a period.
Average Deal Size
The mean value of closed-won opportunities over a period. An input to pipeline velocity and capacity planning.
Expansion Revenue
Additional recurring revenue from existing customers: more seats, higher tiers, additional products or increased usage.
LTV:CAC Ratio
Lifetime value divided by acquisition cost. Whether a customer returns more than they cost to win.
Segmentation
Dividing the market into groups that warrant different treatment — different motion, pricing, coverage or support model.
COMMON QUESTIONS
- What does ARPA mean?
- Average revenue per account: recurring revenue for a period divided by the number of active accounts in it. It expresses roughly what a customer is worth per period, and its direction indicates whether the business is moving up-market or down. The definitional work is in the denominator — accounts, logos and billing entities give different answers.
- What is the difference between ARPA and average deal size?
- Average deal size describes what you sell at the moment of sale; ARPA describes what an account is worth on an ongoing basis, including any expansion since. They diverge in any business with meaningful expansion, and that divergence is informative: ARPA materially above average deal size means the land-and-expand motion is working.
- Is a rising ARPA always good?
- No, and this is the most common misreading. ARPA rises when you win larger customers and also when you lose smaller ones, and those are opposite outcomes with identical trend lines. Always report ARPA alongside account count: rising ARPA with falling accounts is churn at the small end being reported as success.
- Should ARPA use mean or median?
- ARPA is conventionally a mean, which makes it sensitive to a handful of very large accounts. It is worth computing the median periodically as a sanity check — in most businesses the median account is materially smaller than the mean, and knowing the size of that gap tells you how much of your revenue depends on a small number of relationships.
Learn how to apply this: RevOps 101: Revenue Operations Foundations
Definitions are the vocabulary. The courses are where you learn to operate it, with the interactive audit tools.
See the course