Usage-based consumption revenue and the constraint on variable consideration

Many SaaS platforms bill on metered usage (API calls, gigabytes processed, transactions executed) rather than fixed subscription fees, so revenue must often be estimated before actual usage for the period is fully known, and the IFRS 15 constraint exists precisely to stop companies from front-loading optimistic usage estimates into current-period revenue. Because usage-based pricing is increasingly common as SaaS companies shift toward consumption models, getting the estimation and constraint judgment wrong creates a recurring risk of revenue being reversed in a later period once actual usage comes in lower than estimated.

Does the usage-based fee qualify for the series exception, where the contract is a single performance obligation made up of distinct time increments? If the series exception applies, is period revenue measured using the right-to-invoice practical expedient based on actual usage the entity has a right to bill? For usage not yet fully known at period end, how reliable is the estimate (e.g., system metering data) used to avoid overstating or understating revenue?

1. Determine whether the usage-based fee qualifies for the variable consideration allocation exception (the 'series' guidance): this applies when the contract is a series of distinct time increments (e.g., each day or month of access) accounted for as a single performance obligation, and the variable fee relates specifically to outcomes of that period and is allocated entirely to it.
2. If the series exception applies, recognize revenue as usage actually occurs each period using the amount to which the entity has a right to invoice, generally without needing to estimate and constrain the full contract's variable consideration upfront.
3. If the series exception does not apply (e.g., usage-based fees are part of a broader, non-distinct combined obligation with fixed fees, or the contract structure does not meet the series criteria), estimate the variable consideration using the expected value or most likely amount method and apply the constraint (include only the amount highly probable not to result in a significant future reversal).
4. At each period end before the usage report/invoice is finalized, estimate unbilled usage using the best available data (system metering, historical patterns, partial-period actuals extrapolated) rather than waiting for the next billing cycle.
5. True up the estimate to actual usage in the following period as data becomes available, treating differences as a change in estimate in the current period, not a prior-period error, provided the original estimate was reasonable given information available at the time.

For series-exception contracts, measure period revenue as the amount the entity has the contractual right to invoice for that period's actual usage (the 'right to invoice' practical expedient), which closely approximates the value transferred and avoids estimating a full contract-value figure. For non-series contracts requiring full estimation, measure variable consideration at the probability-weighted expected value (for a range of possible usage outcomes) or the single most likely amount (for essentially binary outcomes), then apply the constraint before including it in the transaction price. Where period-end usage data is incomplete, estimate using extrapolation from available metering data plus historical seasonal patterns, refreshing the estimate as more actual data becomes available within the period-end close process.

Disclose: the nature of variable and usage-based fee arrangements and their proportion of total revenue where significant; whether the series exception/right-to-invoice practical expedient is applied and to which revenue streams; the policy and methodology for estimating unbilled usage at period end, including reliance on system-generated metering data; significant changes in variable consideration estimates and their effect on revenue during the period; the constraint methodology applied for any non-series usage-based arrangements.

• Recognizing usage-based revenue only when invoiced, ignoring the need to accrue unbilled usage at period end
• applying the series/right-to-invoice expedient automatically to all usage-based fees without verifying the contract actually meets the series criteria (a single combined obligation of distinct time increments)
• failing to apply the constraint to usage estimates for genuinely uncertain, new, or highly variable consumption patterns, resulting in optimistic overstatement
• using a flat historical average without adjusting for known changes (e.g., a large customer'
• s known usage ramp-down) that make historical data a poor predictor
• not truing up estimates promptly once actual metering data diverges materially from the period-end estimate.