Usage-based pricing

How is usage-based pricing analysed under IFRS 15?

There are two ways to read a usage contract. In the first, the company promises to stand ready to provide its platform and the customer agrees to pay for whatever it uses: the usage fees are variable consideration. Because each fee relates to service in a specific period, IFRS 15 allows it to be allocated entirely to that period, so revenue follows usage. In the second, the customer has the right, but no obligation, to buy more units of a distinct service: each use is an optional purchase that becomes a contract when made, unless the option gives the customer a material right. Both readings usually lead to revenue as the customer uses the service. The distinction still matters for estimates at the period end, the disclosure of remaining performance obligations and material rights. Where the amount billed corresponds directly to the value delivered, the right to invoice practical expedient lets the company recognise what it can invoice. See variable consideration.

Usage-based pricing with prepaid credits

A customer prepays US$ 1,200 thousand for non-refundable credits that can be used over twelve months; unused credits expire. Based on experience with similar customers, the company expects 1,000 thousand of credits to be used, leaving breakage of 200 thousand. IFRS 15 recognises expected breakage in proportion to the pattern of rights the customer exercises, so each unit used carries 1.2 times its face value in revenue.

Breakage is included only to the extent it is highly probable that recognising it will not lead to a significant reversal. If the company could not estimate it reliably, it would recognise revenue only as credits are used, 1,000 thousand over the year, and the remaining 200 thousand when the credits expire or the chance of their use becomes remote. Estimates are updated each period, with a cumulative catch-up. Unclaimed property laws in some places require unused balances to be paid to the government, in which case that part is never revenue. The same logic applies to prepaid airtime and gift cards.

How are minimum commitments and overages treated?

Many enterprise contracts combine a minimum annual commitment with fees for usage above it. How the minimum is recognised depends on what the company has promised. If it promises a quantity of usage, the minimum is recognised as that usage happens, with any expected shortfall treated like breakage and recognised in proportion or when the year ends. If it promises to stand ready to provide the platform for the year, with usage priced on top, the minimum is recognised evenly over the year and only the excess usage as it occurs. A shortfall payment at the end of the year is part of the fixed minimum, not extra revenue. Overage fees are variable consideration recognised in the period of use.

What about tiered and volume pricing?

With incremental tiers that reset each period, for example the first million API calls a month at one rate and the rest at a lower rate, each call is recognised at its tier price in the month it occurs. Retrospective volume discounts are different: if all calls in the year are repriced once annual usage passes a threshold, the company estimates the expected average price for the year and recognises revenue at that rate from the start, with a liability for the expected credit. A lower price on future usage earned by past purchases can be a material right if the customer would not otherwise get it, so part of the earlier fees is deferred.

How are usage estimates made?

For most pay-as-you-go contracts, no estimate is needed beyond usage that has happened but not yet been billed at the period end, which is accrued. Estimates are needed for breakage, retrospective tiers and minimums with expected shortfalls; companies use the expected value or most likely amount, apply the constraint and base them on historical usage, seasonality and customer cohorts. Because revenue depends on metering data, the completeness and accuracy of usage records and the controls over them are key areas for auditors.

What about licensed software priced by usage?

For a licence of intellectual property, IFRS 15 has a specific exception: sales-based or usage-based royalties are recognised only when the later of the usage occurring and the licence obligation being satisfied happens. So on-premise software licensed per transaction is recognised as transactions occur, without estimating future usage. See software licence or SaaS and IFRS 15 licences.

What is disclosed, and how does US GAAP differ?

Companies often disaggregate revenue between subscription and usage, and explain judgements on breakage and minimums. In the disclosure of remaining performance obligations, IFRS 15 allows usage fees to be left out only through the right to invoice or one-year expedients. ASC 606 also allows variable consideration allocated entirely to a distinct period of a series to be excluded, so US companies often disclose smaller remaining obligations for usage-heavy contracts. Otherwise the two frameworks give the same answers. How usage feeds into ARR and other SaaS metrics varies by company. See SaaS accounting.

This guide is general information. It is not tax or legal advice for your situation.