If you are moving your SaaS product from seat-based to usage-based pricing, some of your key metrics will look worse even if the business is getting stronger. Investors will notice before you explain it, so the explanation needs to be ready first.
The reason is mechanical. Usage-based models produce lower bookings and deferred revenue than annual upfront contracts, even when customers are just as committed. As AI products push more software companies towards consumption pricing, more balance sheets will show this pattern.
An experienced analyst knows how to read those numbers. A VC associate reviewing your data room, or a buyer’s due diligence team, may not give you that benefit of the doubt. For UK founders raising a round or preparing for sale in the next 18 months, this is worth getting ahead of.
Why so many SaaS companies are switching
AI is the main driver. When a product’s value comes from work an AI agent does, charging per seat stops making sense. If your software lets one person do the work of three, a per-user price shrinks exactly as you deliver more value.
Cost is the second driver. AI features carry real inference costs that rise with usage. A flat seat price means your heaviest users can quietly erode your gross margin. Charging per task, per credit or per outcome keeps revenue in step with cost.
The result is that many companies are moving to hybrid models: a platform fee plus usage, prepaid credits, or pure pay-as-you-go. Each is a sensible commercial decision. Each also changes how your financials read.
Same revenue, very different numbers
Here is a simplified illustration. Two SaaS businesses each earn £100,000 of revenue a month, £1.2m a year. Company A sells annual contracts invoiced upfront on 1 January. Company B charges for usage, invoiced monthly in arrears on 30-day terms. Usage is flat, so both recognise identical revenue.
| Position at 31 March | Company A (annual upfront) | Company B (usage in arrears) |
|---|---|---|
| Revenue recognised to date | £300,000 | £300,000 |
| Deferred revenue on the balance sheet | £900,000 | £0 |
| Cash collected | £1,200,000 | £200,000 |
| Trade receivables | £0 | £100,000 |
| Contracted revenue for the rest of the year | £900,000 | £0 |
The businesses have the same revenue and the same customers. Yet Company B has no deferred revenue, a sixth of the cash, and nothing contracted beyond the current month. To an investor skimming the balance sheet, Company B looks riskier. It may not be, if its customers are sticky and usage is growing. But it has to prove that, rather than having the contract prove it for them.
The numbers are illustrative and ignore VAT, churn and payment delays.
The six metrics that change
Deferred revenue. Usage billed in arrears creates no deferred revenue at all. Investors often treat deferred revenue as a signal of contracted future income, so a falling balance can read as weakening demand. The exception is prepaid credits, which do create deferred revenue until they are consumed.
ARR. Annual recurring revenue assumes revenue is contracted and repeating. Usage revenue is neither, strictly. Many companies annualise recent usage and call it ARR, but definitions vary widely. An inflated or unclear ARR figure is one of the fastest ways to lose credibility in diligence.
Net revenue retention. NRR becomes more volatile, because it now moves with customer activity, not just contract renewals. It can look spectacular in a strong quarter and worrying in a quiet one.
Cash flow and working capital. Moving from annual upfront to monthly in arrears can take months of cash out of the business at the point of transition. Growth then consumes cash rather than generating it. This catches founders out more than any other change, and it matters when you plan your runway.
Gross margin. If AI inference costs sit in cost of sales, gross margin can move with product mix and usage patterns. Investors will want to see margin by revenue stream, not one blended figure.
Forecasting. Without contracted revenue, your forecast depends on usage assumptions. Investors will test those assumptions harder than they would test a renewal schedule.
What investors and buyers will ask
If you have moved to consumption or hybrid pricing, expect these questions from a lead VC or an acquirer’s diligence team:
- How much of your revenue is committed, and how much depends on usage?
- How exactly do you calculate ARR, and has the definition changed?
- Why has deferred revenue fallen while revenue has grown?
- How does usage behave by customer cohort over time?
- What drives usage spikes and dips, and how concentrated is usage in your largest customers?
- What is gross margin on usage revenue once inference costs are included?
- What happened to cash conversion when you changed pricing?
None of these questions is hostile. But slow or inconsistent answers are where valuations get chipped. In my experience, buyers treat an unexplained change in metrics as a risk, and they price risk.
How to present it: a practical playbook
- Split your revenue into committed and variable. Report platform fees and minimum commitments separately from usage above those minimums. This one change answers half the diligence questions before they are asked.
- Define your run-rate metric and stick to it. If you annualise usage, say exactly how: for example, the last three months of usage revenue multiplied by four. Label it as usage run-rate, not ARR, and never change the method quietly between reporting periods.
- Show usage by cohort. A chart of how each customer cohort’s usage grows over time is the best evidence that variable revenue is actually durable. It does the job a renewal schedule used to do.
- Report gross margin by revenue stream. Show platform and usage margins separately, with AI inference costs clearly allocated. If usage margin is improving as you optimise models, that is a strong story to tell.
- Bridge the metrics that moved. Prepare a one-page reconciliation explaining why deferred revenue, billings or cash conversion changed at the point of the pricing switch. Put it in the data room before anyone asks.
- Protect your cash. Consider annual prepaid commitments or credit bundles for larger customers. They restore some upfront cash and deferred revenue without abandoning usage pricing.
- Stress-test your forecast. Model downside usage scenarios and show investors you have. A forecast that only works if usage keeps rising will not survive a diligence call.
One technical point: the revenue recognition for usage, credits and minimum commitments under FRS 102 or IFRS 15 needs care, particularly where credits expire unused. Get it right before your numbers go in front of an investor, because restating later is far more damaging than explaining now.
The bottom line
Usage-based pricing is often the right commercial call for an AI-enabled product. The problem is rarely the model itself. It is metrics that change shape without an explanation ready, just as an investor or buyer starts looking.
The founders who handle this well do three things. They separate committed from variable revenue, define their metrics honestly and consistently, and explain every movement before they are asked.
If you are moving to consumption pricing, or planning a raise or exit after already making the switch, I am happy to look at how your numbers will read to an investor. Book a free 30-minute strategy call and we can walk through it.
Kish Patel is an ICAEW Chartered Accountant and founder of Consult EFC, a fractional CFO and corporate finance practice for UK SaaS founders.
Not sure where your business stands right now?
Book a free 30-minute call with Kish. Bring your numbers, your questions, or just your situation. You will leave with a clearer picture than you arrived with.
Book a Free Strategy Call
Over 12 years across Big Four audit, Investment Banking, and corporate advisory. Kish works with SaaS founders, tech companies, and ambitious UK SMEs from £1M to £50M in revenue on fundraising, valuations, exit planning, and financial strategy. ICAEW regulated. Big Four trained. Based in London.