ARR: What Annual Recurring Revenue Measures and What It Leaves Out

Annual recurring revenue, almost always written ARR, is the annualised value of the recurring subscription contracts a business has in force at a particular moment. It answers the question: if nothing changed from today, how much subscription revenue would this business collect over the next twelve months? It is a run rate rather than an accounting figure, it is measured on a date rather than across a period, and it deliberately excludes anything that does not repeat by default, such as setup fees, training, hardware and one-off professional services. Because it strips out the noise of when money happened to arrive, it is the cleanest single number for comparing the size and growth of subscription businesses.

There are two routes to the same figure, and which one you use depends on how your customers pay.

From monthly recurring revenue

Add up the recurring subscription value of every live customer for one month, then multiply by twelve. Live means the subscription is active on the measurement date, not that an invoice happened to be raised that month. Recurring means the charge repeats by default without a new purchase decision.

From annual contracts directly

Add the annual subscription value of every live contract. Where a contract runs for more than a year, divide the total committed value by the number of years so that only one year is represented. A three-year contract worth ₹36,00,000 contributes ₹12,00,000, not the full amount.

A worked example (illustrative figures)

A business has 120 customers paying an average of ₹9,000 a month, giving monthly recurring revenue of ₹10,80,000 and an annualised figure of ₹1,29,60,000. It also signs one three-year contract worth ₹36,00,000 in total, which adds ₹12,00,000 to the annual figure, taking it to ₹1,41,60,000. During the same month it invoices ₹6,00,000 of implementation fees. Those fees appear in revenue for the year, but they add nothing to the recurring figure, because no part of them repeats. Anyone comparing the two numbers without that explanation will conclude the reporting is wrong when it is simply measuring two different things.

The recurring figure exists to make a subscription business legible. Monthly revenue in a subscription company is distorted by billing timing, annual prepayments and one-off work, so it is a poor guide to underlying size. Annualising the recurring base removes those distortions and produces a number that can be compared across months, across competitors and across funding rounds.

Operationally, it drives planning. Headcount, hiring pace and spending are usually set against the recurring base rather than against cash collected, because the base is the part of revenue you can reasonably expect to still be there next quarter. It also frames growth conversations properly: the movement in the recurring figure between two dates can be decomposed into new business, expansion, contraction and churn, and that bridge tells you far more about the health of a company than the headline growth percentage does.

The definitional traps

The first trap is including revenue that does not recur. Implementation fees, custom development and training days all feel like real money, and they are, but folding them in means the figure no longer predicts anything. The second is annualising a good month: taking whatever was invoiced in a strong month and multiplying by twelve produces a number that includes one-off items and excludes customers who paid annually in a different month. The third is double counting multi-year contracts by recognising the whole committed value in the current year.

The version that flatters the business

Because no accounting standard governs the figure, it is unusually easy to widen quietly. Counting signed but not yet live contracts, including trials that have not converted, keeping churned customers in the base until their formal contract end date, and treating optimistic usage revenue as committed all push the number up without anything changing commercially. Each individual choice can be argued for. The problem is cumulative: a figure adjusted in four flattering directions is no longer comparable with anyone else's, or with the same company's own figure from a year earlier. Writing the definition down and restating history when it changes is the only durable defence.

The total on its own says less than its composition. Two businesses with identical recurring bases can be in completely different positions depending on how much of it sits in a handful of large accounts, how much comes up for renewal in the next two quarters, and how much of the growth came from expansion rather than new logos. Ask for the bridge before the total: opening base, new business, expansion, contraction, churn, closing base.

Concentration is the risk most often missed. A recurring base where a small number of customers represent a large share of the total behaves very differently in a downturn from one spread across many small accounts, even though the headline figure is the same. Renewal timing matters for the same reason. A base with most of its contracts renewing in one quarter carries a cliff that an evenly distributed book does not.

These four figures describe the same contracts and answer different questions.

Monthly recurring revenue belongs in the same family and differs only in scale. Committed recurring revenue is a fifth variant that adds signed but not yet live contracts, which is helpful for businesses with long onboarding and misleading if it is not labelled.

A trustworthy recurring figure depends on clean contract records rather than clever spreadsheets. Start with what a CRM is, then look at lead management software, sales automation for renewal cadences, CRM versus Excel for why overwritten fields destroy the audit trail, product features, and CRM for small business if you are putting this in place for the first time.