Under the bonnet

How the matching engine works

What the ranking is built from, what the percentage means, and the rules the scoring is never allowed to break.

The short version

You describe the deal. The engine checks it against every lender on the panel — their published criteria, what they have told us directly, and what they have actually done on deals shaped like yours — and returns a ranked list with a probability of terms against each name. Not a directory. Not twenty phone numbers. A ranking that says who wants this deal, this month, and why.

Three layers of evidence

Hard criteria first. Loan size, borrowing level, asset type, location, charge position. These only rule lenders out. A lender who caps at 70% does not appear at 80% — nothing clever about it, and nothing hidden either: every exclusion carries its reason.

Live appetite second. Criteria sheets date quickly. Lenders drift, funding lines fill, sectors fall out of favour. The engine weighs what each lender has been doing recently — the deals they quoted, the ones they declined and the reasons they gave — so the ranking reflects this month, not January's rate card.

Judgement third. Twenty years of placing bridges leaves patterns no spreadsheet holds: who genuinely does what they advertise, who slows down above a certain size, who reads a planning story properly. That knowledge is written into the engine as rules, reviewed and signed off by a human every time it changes.

What the score means — and what it doesn't

The percentage against each lender is a probability of terms, not a promise. Where the evidence is thin, the score is capped and says so. Where our information on a lender is old, the ranking marks it as unverified rather than pretending. If nothing fits, the honest answer is nothing fits — you will see that too, with the detail that most often changes it.

Two more rules the scoring lives by. Commission never moves a fit score: lenders cannot pay to rank higher, and where a commercial arrangement affects routing priority it is shown, separately and labelled. And every score explains itself in plain English — headroom, exit strength, the things a lender will question — because a number you cannot interrogate is a number you should not trust.

Why it gets better every week

Every real outcome feeds back in. A quote, a decline with its stated reason, the time a lender took to answer — each one adjusts what the engine believes about that lender's appetite. The panel is measured by behaviour, which is the one thing that cannot be faked in a rate card.

The engine reads, checks and ranks. It does not make lending decisions, and it does not replace the specialist — every application is reviewed by a human before it reaches a lender.
Want this worked through on your numbers?

Tell the engine your version in plain language. It asks what an underwriter would ask, then shows which of 300 tracked lenders actually want your deal — ranked, with honest probabilities.

Get matched →

Four costs matter on almost every bridge: the monthly interest rate, the arrangement fee, the valuation and the legals. As of July 2026, unregulated bridging under 75% loan-to-value prices at roughly 0.79%–1.20% per month across our panel. The arrangement fee is typically 2% of the gross loan. Valuations run from a few hundred pounds on a straightforward house to several thousand on commercial. And in bridging, unusually, the borrower normally pays the lender's legal costs as well as their own.

A worked example

Borrow £400,000 for twelve months at 0.9% a month, interest rolled up.

ItemCost
Interest (12 months, rolled)£43,200
Arrangement fee (2%)£8,000
Valuation£600
Legals (both sides)£2,500
Total cost of funds~£54,300

Call it 13–14% for the year, all in. Nobody should pretend that's cheap, and we won't. You pay it for one of two reasons: the deal itself makes far more than the money costs, or the clock is real and missing it costs you the deal entirely. If neither applies, a bridge is the wrong product and a decent broker will tell you so before you spend a pound.

The costs people miss

Exit fees have largely died out on mainstream bridges — most of our panel charge none — but a minority still take 1% or a month's interest on the way out, so check the redemption terms, not just the headline rate. Watch minimum terms too: some lenders charge at least three months' interest even if you repay in six weeks. And if your loan is rolled-up, remember the interest compounds: month nine's interest is charged on a bigger balance than month one's.

What a broker costs on top

Here, nothing. bridgingradar is paid a procuration fee by the lender on completion, the same way the wider market works — the difference is that we show you the exact figure on every deal, disclosed in the paperwork. Nothing is added to your rate to fund it.

Cost the delay, not just the loan. The expensive bridge is rarely the one with the higher rate — it's the one with the optimistic term. Set the term to the plan going wrong by two months, not to the plan going right.

Matthew Dailly — arranging bridges since 2004
Dated numbers: rates and fees above reflect what we see across our tracked lender panel in July 2026. Your figure depends on the asset, the leverage and the exit — the engine will show you lender-specific pricing in minutes.

Related reading

Rolled up, retained or serviced — how is bridging interest charged?How fast can a bridging loan complete, realistically?What deposit do I need — and can additional security replace it?

← Knowledge Base

Your deal. 300 lenders. One decision-ready answer.

Get matched now →
Free · no obligation · no credit footprint