A profitable-looking SaaS business can still be starved of cash, because it pays for a year of customer acquisition up front and collects the subscription back over the following twelve months. The traditional bank sees a company with no plant, no stock and thin or negative profits, and offers very little. Two non-dilutive sources fill that gap, and a SaaS finance director should treat them as distinct instruments with different underwriting. One advances against the recurring revenue itself. The other advances against the R&D tax credit HMRC already owes you. Neither costs a point of equity, which is the entire reason to reach for them before a round.
Lending against recurring revenue, and what actually sets the limit
Recurring-revenue lenders do not underwrite your balance sheet; they underwrite the predictability of your subscription book. The advance is sized against ARR or MRR, and the multiple a lender will offer turns on three numbers that a generic business loan never looks at.
- Net revenue retention. This is the headline. A book that grows inside its existing customer base, where expansion and upsell more than offset churn, supports a larger and cheaper facility than one leaking customers, even at the same ARR. Retention above one hundred percent means the revenue base compounds without new sales, and lenders price that as durable.
- Gross margin. Strong SaaS gross margins, commonly seventy-five percent or higher, mean each pound of revenue throws off enough contribution to service the facility. A business carrying heavy hosting, support or services cost inside its "recurring" line gets sized down once the lender strips out the parts that are not really software margin.
- Churn and contract shape. Annual contracts paid up front read very differently from rolling monthly subscriptions a customer can cancel at any time. Longer committed terms with a creditworthy customer base lift the advance rate.
The mechanics matter. Some facilities advance a large share of contract value up front and collect as customers pay; others run as a revolving line that flexes monthly with bookings, so you draw more as ARR grows and the limit re-sizes off live data rather than a year-old set of accounts. The limit of the product is that it only works where revenue is genuinely recurring and genuinely retained. A business calling project work or one-off implementation fees "recurring" gets repriced the moment the lender reads the cohort detail. Lead with cohort-level retention by start date, not a flat headline number, and the facility comes together faster. Bedrock places these against the subscription book through cashflow loans and, where the contracted invoices to business customers are the cleaner asset, through invoice finance on the B2B receivables. How this sits against the wider funding stack is on the technology sector page.
Advancing the R&D credit before HMRC pays
The second source is the money the taxman already owes you for building the product. For accounting periods beginning on or after 1 April 2024, the previous SME and RDEC reliefs were consolidated into a single merged R&D expenditure credit scheme, an above-the-line credit at a 20 percent headline rate that nets to roughly 15 to 16 pence per pound of qualifying spend after corporation tax. Loss-making, R&D-intensive SMEs sit in a separate, richer regime: Enhanced R&D Intensive Support, available where qualifying R&D is at least 30 percent of total expenditure, which delivers materially more cash per pound spent. A SaaS business writing proprietary software is often exactly the kind of claimant these schemes are designed for.
The problem is timing. The credit only turns into cash when HMRC processes the claim after your accounting period closes, which can be many months after you spent the money on the engineers who earned it. Specialist innovation lenders bridge that gap. They advance against the pending claim, commonly up to around 80 percent of a verified forecast and less for a first-time claimant, then take repayment directly when HMRC pays the credit to them and return the surplus to you. The underwriting question is not your profitability but the quality and certifiability of the claim, which is why a well-prepared claim with a credible adviser behind it borrows more cheaply than a speculative one. Used well, this pulls forward cash you have already earned, by up to a year or more, to fund the next stretch of payroll without diluting anyone.
Stacking them, and the limit of non-dilutive
These two sources are complementary, not competing. The R&D advance is a one-off-per-period bridge against a known sum the state owes you; the ARR facility is an ongoing line that scales with the subscription book. A growing SaaS business can run both, advancing the credit once a year to fund the R&D cycle and drawing on the recurring-revenue line to fund go-to-market between claims. Layering them in that order keeps the blended cost of capital down and the cap table untouched.
The honest limit is that non-dilutive debt funds growth a business can already see, not bets it cannot yet prove. Both instruments size off something concrete: a retained revenue book or a quantified claim. Neither will fund a pivot, an unproven new product line, or the open-ended runway of a business still searching for its market. That is genuine risk, and it is paid for in equity, not interest. The discipline is to exhaust the non-dilutive sources against everything that is already contracted or already earned, and only sell ownership for the part of the plan that is still a bet. SaaS founders who default to a round for working capital they could have borrowed against give away equity they did not need to part with.