For many AI and software-as-a-service businesses, revenue growth can move quickly.
New customers are signed, contracts are renewed, usage expands and product features evolve at pace.
The accounting treatment behind that growth, however, can be less straightforward.
Revenue recognition is a particular challenge for high-growth technology companies because commercial contracts often include several moving parts. A single customer arrangement might combine platform access, implementation support, training, premium support, data migration, bespoke development, usage-based fees or third-party services. Each element may need to be considered carefully before income is recognised.
With changes to FRS 102 taking effect for accounting periods beginning on or after 1 January 2026, now is a sensible time for AI and SaaS businesses to review their contracts, systems and year-end processes. The updated UK GAAP revenue model is more closely aligned with IFRS 15 and places greater emphasis on identifying performance obligations, determining the transaction price and recognising revenue when those obligations are satisfied.
Why SaaS revenue is rarely simple
Traditional product sales can often be easier to account for because there is a clearer point at which control passes to the customer. SaaS businesses usually sell access to a service over time. If a customer pays upfront for a 12-month subscription, the cash may be received on day one, but the revenue will typically be earned over the subscription period as the service is provided.
Complexity increases when the contract includes additional promises. Implementation services may be distinct from the subscription, or they may be so closely linked to the platform that they need to be accounted for together. Training, onboarding and support may also require judgement. The way these promises are identified can affect both the timing and the profile of revenue recognition.
For AI businesses, contracts can be particularly nuanced. Pricing may be based on seats, consumption, API calls, outputs, compute usage, data volume or success-based metrics. There may also be service level commitments, credits, discounts, renewal options or termination rights. These terms can affect the transaction price and whether any element of consideration should be deferred or allocated differently across the contract.
Another area to review is whether the business is acting as principal or agent. Many technology companies bundle third-party cloud hosting, software tools, data services or marketplace products into their customer arrangements. Where a company controls the service before it is transferred to the customer, gross revenue recognition may be appropriate. Where it is merely arranging for another party to provide the service, net presentation may be required.
Revenue recognition is not just a technical accounting issue. It can influence management reporting, bank covenant calculations, investor communications, tax computations and business valuations. For companies seeking investment or preparing for an exit, inconsistent or poorly documented revenue policies can create avoidable due diligence questions.
Year-end planning is therefore an opportunity to check whether revenue policies still reflect how the business sells today. Fast-growing SaaS and AI companies often introduce new pricing models faster than their accounting processes evolve. What started as a simple subscription product may now include enterprise licences, usage tiers, implementation projects, customer success packages and bespoke AI development work.
Practical steps for AI and SaaS businesses
Businesses should start by reviewing a representative sample of customer contracts, including any non-standard enterprise deals. The review should consider what has been promised to the customer, whether each promise is distinct, how the price has been set and when the customer receives the benefit of each service.
Finance teams should also consider whether billing systems, CRM data and management accounts capture the information needed to support the accounting treatment. If revenue needs to be allocated across multiple performance obligations, the business may need reliable standalone selling prices or a clear methodology for estimating them. If usage-based or variable fees are material, controls around estimates and subsequent adjustments will be important.
Documentation is key. Auditors, investors and lenders will expect to see the reasoning behind revenue judgements, particularly where contracts are material, bespoke or unusual. A clear revenue recognition policy, supported by contract analysis and evidence of review, can reduce year-end pressure and help avoid surprises.
At Hillier Hopkins, we work with technology and high-growth businesses to review accounting treatment, assess contract terms and prepare for reporting changes. If your business is reviewing its revenue recognition policies or planning for year end, speak to our team for practical, commercially focused advice.
