Reported ARR Is a Metric. Bankable ARR Is a Verdict.

A young London SaaS company secured debt without profit because it could prove the behaviour of its customer cohorts. That is the clue for rollup buyers: lenders are not financing a dashboard total, but the portion of revenue likely to remain after ownership changes.

A London SaaS compliance management business was only 18 months old. It had £380,000 of annual recurring revenue, or ARR, the annualised value of its subscription income. It was not profitable. Conventional banks turned down its request for an unsecured loan.

Then Doulton Bridging Finance, a UK specialist lender, approved £150,000. The money arrived within seven working days and helped fuel a later Series A. The lender did not treat the company’s youth as the whole story. It examined an ARR waterfall, customer cohorts, churn, customer contracts and a 12-month model. Its reported net revenue retention, or NRR, was 108%: existing customers, taken together, were spending more over time despite any losses.

Doulton’s loan shows how documented cohorts and contracts can matter more than current profitability. Photo: AXP Photography / Pexels, Pexels licence (free commercial use).

That is a small financing, not a completed SaaS rollup. It is also the clearest available UK clue to a much bigger question: when a buyer wants debt to acquire several software companies, what exactly is a lender lending against?

Not the number at the top of the dashboard. Not on its own.

Most founders still approach acquisition finance for SaaS rollups as a scale test. Get enough ARR. Show growth. Apply a familiar leverage multiple, meaning a multiple used to estimate how much debt a company might support. Job done.

That mental model has become dangerous. The lender is really asking a more awkward question: if the business changes hands and the products, teams and billing systems are combined, which subscriptions are still likely to be there?

The number that shrinks in the credit paper

Imagine two software businesses each reporting the same ARR. One has hundreds of customers who renew reliably. The other depends heavily on one customer, has a fraying recent cohort and has offered repeated discounts to keep renewals coming. The dashboard gives them equal status. A lender does not.

Round Rock Requisition, which sets out ARR-loan underwriting criteria, reports that lenders can apply 10-15% haircuts, meaning deliberate discounts to the revenue they will recognise for lending, when monthly churn sits at 2-3%. It also says that a single customer contributing more than 25% of ARR is excluded from the financeable base.

Its published underwriting criteria show why reported ARR can be reduced before a lender sizes debt. Photo: Goszton / Pexels, Pexels licence (free commercial use).

That last rule answers one of the most common questions founders ask: what percentage of revenue comes from the top five customers? There is no universal safe percentage in the evidence here. But there is a hard warning sign. If just one customer exceeds 25%, an ARR lender may count none of that customer’s revenue when sizing the loan.

A company can therefore look bigger than it is to a buyer. Or, more precisely, bigger than it is to a credit committee.

What surprised us was how often the discussion begins with a minimum monthly recurring revenue threshold, then ends somewhere else entirely. There is no minimum MRR in this evidence that unlocks acquisition debt by itself. The relevant threshold is behavioural: can the buyer demonstrate that the income is durable, diversified and transferable?

This is why a neat ARR chart can be less useful than a scruffy but reliable cohort export. A cohort is simply a group of customers that started around the same time. It tells a lender whether customers who joined last year are still paying now, rather than allowing a surge of new sales to hide losses among older customers.

The lender’s version of ARR is smaller. Call it bankable ARR.

GI Network's view: In a SaaS acquisition, the revenue figure that matters is not what customers paid before the sale. It is what a lender can reasonably expect them to keep paying after the buyer starts changing things.

The young company that made proof matter more than profit

The Doulton case overturns a comforting belief and replaces it with a more demanding one. An unprofitable company can borrow. But it must arrive with evidence that makes its revenue intelligible.

Doulton’s London borrower had £380,000 ARR and 108% NRR. Crucially, the credit pack did not stop at the headline. It included the ARR waterfall, cohort data, churn metrics, customer contracts and a forward model. Those documents allowed the lender to inspect the moving parts beneath the total.

The London SaaS case made retention evidence, rather than a long trading history, the centre of the lending decision. Photo: Daniel Case, compositing and London image; Singapore image by XRay, a/k/a Dietmar Rabich. / Wikimedia Commons, CC BY-SA 4.0.

For a rollup buyer, this distinction gets sharper. The buyer is not merely asking a lender to accept that customers renewed under the old owner. It is asking the lender to accept that they will renew through a change of control, a new product roadmap, a possible billing migration and the inevitable distraction of combining companies.

The research brief does not contain a completed UK rollup where a lender publicly quantified an “integration penalty” or disclosed the exact debt reduction it imposed. That matters. It would be easy, and wrong, to invent a clean formula.

Still, the underwriting logic is visible. Financely, an adviser on acquisition financing for software and SaaS businesses, says lenders price and size risk around retention, ARR quality, customer concentration and cash conversion. In plain English, they want to know whether subscription income turns into cash and stays there. Integration risk belongs in that inquiry because integration is one of the few things an acquirer can do that may make formerly sticky customers leave.

This is where most people stop looking. They treat integration as an operating plan for after closing. A lender has to treat it as a risk to the asset supporting the loan before closing.

The sensible response is not theatre. It is proof. A buyer should be able to show customer-level ARR bridges, cohort-retention exports, renewal pipelines, contracts and records of billing or contract amendments. It should identify which customers are on committed contracts and which can leave monthly. If pricing has changed, or discounts have been used to retain customers, those histories need explaining rather than hiding.

These are not magic documents. They do not erase risk. They give the lender a way to judge it.

A second UK lesson: profit is not the only door

A clinical trials B2B SaaS firm in the UK needed £200,000-£250,000 of growth capital. Mainstream lenders rejected it because it was pre-profit. Plutus, a specialist revenue-based finance provider, funded it based on recurring-revenue stickiness, with repayment tied to a share of revenue rather than an equity sale.

Plutus backed a pre-profit clinical trials SaaS company on recurring-revenue stickiness rather than equity dilution. Photo: Clément Proust / Pexels, Pexels licence (free commercial use).

This is not acquisition finance in the conventional buyout sense. But it reveals the same underwriting instinct. Plutus did not need today’s profit to be perfect if the recurring revenue was sufficiently dependable.

Put the Doulton and Plutus cases beside the Round Rock criteria and a pattern appears that none of them says outright: lenders will sometimes accept the absence of profit, but they will not accept the absence of evidence. Small specialist financings have room for judgement. The price of that flexibility is scrutiny of the customers beneath the revenue line.

For founders asking what SaaS buyers are really looking for right now, that is the answer. Not just growth. They want proof that growth was not bought with fragile pricing, one oversized account or customer relationships that disappear when the founder leaves.

That last issue is particularly uncomfortable for founder-led businesses. What happens when you leave? If customer trust, renewals or product knowledge sit mainly with one founder, the buyer has an integration problem before any systems are merged. The evidence in this brief does not supply a universal metric for founder dependence. It does establish the broader test: revenue must be durable enough to underwrite, not merely visible enough to report.

The larger-market reversal

Now comes the twist. Better ARR evidence can preserve access to debt in smaller financings. It may not be enough once the deal moves into institutional private credit, lending by non-bank investment funds and specialist institutions.

Development Corporate reported on 19 May 2026 that Lincoln International, an investment bank and advisory firm, had placed eight software loans in the first four months of that year. None was underwritten on ARR alone. Lenders required EBITDA, a measure of operating earnings before interest, tax and certain accounting charges, and cash-operating profit instead.

That does not mean ARR is dead. It means the underwriting divide is increasingly about deal size and lender type.

At the smaller end, the Doulton and Plutus cases show that verified retention and contracted revenue can stand in for present profitability. At the larger end, the Lincoln data says lenders want proof that the business produces cash after running itself. The same revenue may support one financing structure and fail another.

This is an assumption worth challenging directly: a great SaaS multiple does not guarantee debt capacity. In fact, a buyer who builds an acquisition price around an optimistic revenue multiple can find that the financing market values the target through a much less flattering lens.

There is a historical parallel. In securities financing, lenders apply haircuts to collateral rather than accepting its face value. Infrastructure lenders similarly prefer contracted revenues to theoretical capacity. SaaS lenders are making the same move in a different costume. They are not buying the story that £1 of recurring revenue is always £1 of dependable collateral.

The accounting wrinkle after the celebration

Even buyers who clear the retention test can meet another constraint after a deal is agreed: deferred revenue. This is cash collected for subscriptions that have not yet been delivered in accounting terms.

CVF Fund notes that, after an acquisition, buyers reporting under IFRS may remeasure acquired licence balances. That can reduce the value attributed to those balances. It is a separate kind of haircut, and it can alter how a transaction is structured.

Here is the twist: a customer may have paid in advance, yet the buyer may not be able to treat every pound of that balance as straightforward economic support for the deal. ARR quality and accounting treatment are different questions. Both can narrow the room for debt.

When ARR really can carry the argument

The opposite case deserves its due. The Doulton and Plutus financings show that smaller UK SaaS businesses can obtain non-dilutive capital before profit. Strong retention, visible churn data and durable contracts are not cosmetic enhancements. They can change the result.

Nor does the evidence prove that every acquisition requires EBITDA-based lending. It shows a direction of travel in eight Lincoln placements, not a law of nature for every lender and every deal.

The practical lesson is more useful than either extreme. Do not tell yourself ARR no longer matters. Do not tell yourself it is enough. Match the capital request to the evidence you can genuinely produce.

Build the pack before you pick the target

For buyers, the work begins before lender outreach, not after a term sheet arrives.

First, produce a bankable ARR bridge for each target. Start with reported ARR. Identify revenue exposed to a very large customer. Show churn by cohort. Separate committed contracts from subscriptions that can end quickly. Explain renewal timing and any pricing or discount changes. This is the same basic discipline behind our examination of why your biggest contract may be the reason you cannot borrow.

Second, treat integration as a customer-retention plan. Which product, support or billing changes could touch customers? Which accounts need named relationship coverage? Which systems are being migrated, and when? The brief does not give a lender-approved template, so do not pretend there is one. Give the lender a credible map of what will change, who owns each risk and how churn will be watched.

Third, test the buyer’s own runway. What does the combined business look like 18-24 months after funding? A lender seeking cash-operating profit will care less about the acquisition announcement than the point at which interest, integration costs and normal operations all have to coexist.

Finally, do not mistake a nearly complete data room for a financeable one. The missing five per cent is often the customer-level trail that turns an attractive narrative into an underwritable case, as we explored in the data room was 95% done. The lenders still would not fund..

What experienced investors see first

Investors are not being fussy when they interrogate concentration, renewal data and integration plans. They are trying to protect against a particular trap: debt makes a disappointing acquisition more consequential.

Equity investors can live with a company missing a forecast if the long-term case survives. A lender has scheduled repayments. It therefore asks whether the recurring revenue is sufficiently predictable to service them, especially when the buyer has taken on the additional job of integrating a target.

Experienced investors also recognise that a customer contract can be both an asset and a vulnerability. A large contract raises ARR today, but can lower financeable ARR if its loss would destabilise repayments. A discount may save a renewal, but may also reveal that list-price revenue is less durable than it appears. None of this makes the company bad. It changes what kind of capital fits.

That is why the right question is not, “What is my micro-SaaS worth?” Value and debt capacity overlap, but they are not identical. A buyer may pay for strategic fit, product capability or cross-sell potential. A lender is paid back by cash that arrives after the deal.

GI Network would begin by rebuilding the reported ARR into a lender-ready bridge, testing concentration, cohort retention, contractual durability, cash conversion and the points at which integration could trigger churn. We would then align the acquisition case, operating plan and supporting documents with the capital providers whose underwriting standards fit the deal, and rehearse the objections that will surface in the credit process. If you are preparing a rollup, apply for capital only after that evidence pack can survive scrutiny.

Use the Five-Subtract Test

Before you ask how much debt a SaaS acquisition can support, run the Five-Subtract Test:

1. 1.Start with reported ARR. This is the dashboard number, not the answer.
2. 2.Subtract concentration exposure. Flag revenue tied to any customer large enough to change the credit case.
3. 3.Subtract renewal fragility. Use cohort churn and upcoming renewals to identify income that may not persist.
4. 4.Subtract pricing fragility. Examine discounts, amendments and any evidence that customers renew only on altered terms.
5. 5.Subtract change-of-owner risk. Identify the revenue exposed to founder departure, product migration, billing changes or integration disruption.

What remains is not a precise lending formula. The research does not provide one. It is the right mental model: reported ARR minus the reasons it may not survive equals the revenue a lender can begin to believe.

That is where acquisition finance starts.