What is SEO for mortgage brokers and specialist lenders?

SEO for mortgage brokers and specialist lenders is the work of making the cases you can place, and the criteria behind them, easy for Google and AI engines to find and quote. For specialist firms, that means publishing plain, dated pages for case types such as bridging, commercial and adverse credit, instead of relying on rate tables and lead forms.

Brokers already handle most of the market. Of the 1.08 million new mortgage transactions in the FCA’s product sales data for 2025, 83% were intermediated, close to 84% the year before (FCA product sales data, reported by Mortgage Solutions, 20 August 2026). The same FCA release recorded 4,499 mortgage advice firms sharing £1.6 billion of broking revenue. That is a crowded field for a search like “mortgage broker near me”.

The competition is also shifting. In July 2025 the FCA removed the rule that made almost any interactive conversation count as advice, which makes execution-only sales easier for lenders to offer directly (FCA, PS25/11, 22 July 2025). The FCA said the change would let intermediaries add value by focusing on consumers with more complex needs. Complex cases are where a specialist broker earns its fee, and they are also where borrowers do the most research before they call anyone.

More of that research now happens in AI assistants. McKinsey’s analysis found that a brand’s own sites account for only 5 to 10% of the sources AI search references (McKinsey, 16 October 2025). What a borrower is told about bridging or a mortgage after a CCJ depends mostly on other sites, which is why the facts on your own pages need to be specific enough to be worth quoting.

Why are shared lead-generation sites an expensive way to win complex cases?

Because a shared lead is, by definition, sold to more than one firm, so brokers compete on how fast they call rather than how well they fit the case. The cost also stops buying anything the day the spending stops. Complex cases suffer most, because a generic enquiry form rarely captures the detail that decides whether a case can be placed at all.

  • Speed beats fit. When the same enquiry reaches several brokers, the first to call usually wins the conversation, whether or not they know the lender panel for that case.
  • The form loses the detail. Loan amount and postcode fit in a form. The date a CCJ was satisfied, the exit route on a bridge or two years of contractor day rates usually don’t.
  • Nothing compounds. Every lead is a one-off purchase. A page that answers “bridging loan to buy before selling” keeps bringing enquiries after it is published.
  • Your firm stays invisible. Borrowers and AI engines see the lead site’s brand, not yours. When an assistant is asked which firms handle a type of case, it has little about you to work with.

Moving off shared leads works best as a gradual shift rather than a switch. Keep paying for volume where you need it, build pages for the two or three case types you know best, and record the source of every completed case. When owned pages start producing enquiries you can measure, move budget across in steps.

Why do broker-only criteria PDFs hide specialist cases from AI engines?

Because the facts that decide a complex case, such as maximum loan to value, accepted credit history, minimum trading period or exit strategy, usually sit in lender criteria guides written for intermediaries, often as PDFs or behind a broker login. AI crawlers cannot log in, and dense PDFs give them little to quote, so borrowers asking AI rarely see these facts.

Demand for these cases is real and rising. FCA data obtained through a freedom of information request showed 10,089 mortgages sold to borrowers with impaired credit histories in 2025, up 23% from 8,183 in 2024 and the first year above 10,000 since 2008 (Mortgage Solutions, 28 September 2026). Those figures exclude product transfers, second charges, further advances and business loans, so the full specialist market is larger.

Each of those borrowers started with a question such as “can I get a mortgage with a satisfied CCJ?” or “how long after a default can I remortgage?”. The honest answer depends on criteria that change between lenders and over time. When those criteria exist only inside broker portals, AI engines answer from whatever general article they can find, often a national comparison or money-advice page with no view of your panel.

The fix is not to republish lenders’ documents. Broker-only guides are often covered by your lender or network agreements, so check those terms before reusing any of that material. What you can publish is your own summary of the cases you place, written in your words, dated, and clearly labelled as a guide rather than a lender’s official criteria.

What should a criteria page show for each specialist case type?

Each case type needs its own page with a short, dated criteria table in plain HTML: who the case suits, what lenders on your panel typically accept, what evidence is needed and whether the loan is regulated. The table below shows what borrowers ask for each case type, where the facts usually sit today, and what a broker can publish instead.

Case typeWhat borrowers askWhere the facts sit todayWhat to publish, with an “as at” date
Regulated bridging (chain break on the borrower’s home)“Can I buy before I sell?”Lender criteria guides, often PDFMaximum term, typical loan-to-value range, exit evidence needed, and a plain statement that the loan is regulated
Investment and commercial bridging“Bridging for an auction purchase in 28 days”Broker portals and case-by-case lender appetitePurposes you place, typical completion times, loan-to-value range, exit routes lenders accept
Owner-occupied commercial mortgage“Mortgage to buy my business premises”Criteria PDFs and lender relationship managersTrading history usually required, typical deposit, property types placed, accounts and documents needed
Semi-commercial and mixed use“Mortgage for a shop with a flat above”Criteria PDFs, split by lenderHow the residential share affects regulation, property types placed, typical deposit
Adverse credit“Mortgage with a satisfied CCJ”Tiered lender criteria, often broker-onlyCredit events considered, how recency and amount matter, evidence a lender will ask for
Self-employed and contractors“Mortgage with one year’s accounts”Income sections of lender guidesTrading periods your panel accepts, documents needed, how income is assessed

None of these rows quotes a rate or a fee, and that is deliberate. A criteria page that sets out eligibility, evidence and process answers the questions borrowers ask first, without the figures that bring in the representative-example rules covered in the next section. Where you do want to show costs, put them on a page built to meet those rules.

How do FCA mortgage promotion rules apply to a broker’s web pages?

Any page promoting a regulated mortgage must be fair, clear and not misleading under MCOB 3A.2. If it quotes an interest rate or any cost figure for an MCD regulated mortgage, MCOB 3A.5 requires standard information, including the APRC, set out as a representative example. Pages without rates or costs avoid that trigger, but still need the same care and sign-off.

  • Fair, clear and not misleading. The FCA’s MCOB 3A.2 (last updated 28 February 2025) applies this standard to communications and financial promotions for regulated mortgages. Page titles, meta descriptions and FAQ answers are part of the promotion, because they are what search engines and AI assistants show.
  • Representative examples. Under MCOB 3A.5, an example only counts as representative if the firm reasonably expects at least 51% of consumers who respond and take out the mortgage to be charged the stated APRC or below. The same section recognises that a firm may show a table of several products for comparison, alongside the required overall representative example.
  • Which loans are regulated. The FCA’s PERG 4.4 guidance explains that a regulated mortgage contract is secured on UK land at least 40% of which is used, or intended to be used, as a dwelling, subject to exclusions such as loans to commercial borrowers. Bridging on a borrower’s own home and most investment bridging can therefore sit under different rules, and each page should say which applies.
  • Network sign-off. If you are an appointed representative, your principal firm is responsible for your promotions, so its approval process covers web pages and criteria tables as well as adverts.

How should a broker publish case criteria that AI engines can read?

Start with the case types you place most often and understand best. Give each its own page, open with a direct answer, add a dated criteria table, state whether the loan is regulated and name the adviser who handles those cases. Then check that your firm’s details match the FCA Register and every directory or network listing where you appear.

  1. One page per case type, not per town. “Bridging to buy before you sell” and “mortgages after a CCJ” are separate questions. Hundreds of town pages with the same text answer neither.

  2. Lead with the answer. Open each page with two or three sentences that answer the borrower’s question directly, including the honest limits, before any form or call to action.

  3. Date every criteria table. Show an “as at” date and review the table monthly, or sooner when your panel changes criteria. Remove a row rather than leave it wrong.

  4. Name the adviser. State who handles these cases, their relevant qualifications, your FCA firm reference number and, if you are an appointed representative, your principal firm.

  5. Show anonymised case studies. With the client’s permission, describe the situation, the obstacle and how the case was placed, without rates unless the page meets the representative-example rules.

  6. Match your details everywhere. Firm name, trading names, address and reference number should match the FCA Register, your network’s listing, review sites and business profiles. FinancialService markup can restate those facts, as long as every value matches the page.

How much competitive advantage does SEO give a mortgage broker?

On the Cllimber Opportunity Index 2026, SEO scores 46.3 out of 100 for mortgage brokers, inside the 43 to 49 band where the edge grows the longer a firm keeps at it. Lead generation scores 48.3, CRM 49.8 and sales automation 52.8, the only category in the 50-plus band. The sector average across the six categories is 48.0.

46.3
SEO score out of 100 for mortgage brokers: a strong edge that grows the longer it is sustained
52.8
Sales automation, the highest-scoring category
49.8
CRM, which supports product transfers and repeat business
48.3
Lead generation, just above SEO
45.8
Marketing, just below SEO

Source: Cllimber Opportunity Index 2026, mortgage brokers breakdown (updated 3 July 2026). Social media scores 44.8. Bands: 50+ is a big, lasting edge; 43 to 49 is a strong edge that grows the longer it’s sustained.

The Index estimates the advantage a business gains from using a category of tool well, compared with a direct competitor that doesn’t, across 63 industries. It scores categories, not suppliers or agencies.

Lead generation and SEO sit close together because they solve the same problem from opposite ends. Paid leads deliver enquiries now and stop when the budget does. Search visibility is slower to build but belongs to the firm, and for complex cases it brings borrowers who have already read what you place and why. Most brokers will run both for a while; the scores suggest neither should be ignored.

What should mortgage brokers avoid in SEO and AI search?

Avoid anything that makes borrowing look easier, cheaper or more certain than it is. The common mistakes are headline rates without a representative example, criteria copied once and never updated, near-identical town pages, and promises such as guaranteed approval. Each one creates conduct risk and gives AI engines something wrong to repeat to borrowers.

  • Rates without the required information. A headline rate or monthly cost on a regulated mortgage page brings in the representative-example rules in MCOB 3A.5.
  • Certainty you can’t offer. “Guaranteed approval” or “no credit checks” is not something a broker can promise, and an AI answer that repeats it will mislead a borrower.
  • Criteria pages left to go stale. Lenders change criteria often. A table that was right last spring and carries no date is the one an assistant quotes after it stops being true.
  • Town pages at scale. Google’s spam policies treat pages produced in bulk mainly to rank as scaled content abuse, and since 15 May 2026 they also cover attempts to manipulate generative AI responses.
  • Mixing regulated and unregulated loans. A single page covering home bridging and commercial bridging without saying which is regulated leaves borrowers, and engines, to guess.
  • Paying for promised AI placement. No one controls which sources an AI engine cites. Accurate pages and consistent facts make citation more likely; nothing makes it certain.