What is SEO for financial services?

SEO for financial services is the work of making a regulated firm’s pages easy to find, read and trust, both for Google’s ranking systems and for the AI engines that now answer money questions directly. It differs from ordinary SEO in one way that shapes everything else: the same page is judged by a search engine’s trust standards and by the FCA’s financial promotion rules.

Most of the standard toolkit still applies: crawlable pages, clear headings, fast loading, sensible internal links. What changes is the bar for content. Google treats topics that can affect someone’s financial stability as “Your Money or Your Life” (YMYL) topics, and the FCA’s rules cover what a firm says on its website, its blog and its social channels.

A third reader has arrived. When someone asks ChatGPT, Perplexity, Copilot or Google’s AI Overview which pension provider, lender or wealth manager to look at, the engine builds its answer from sources it has already read. Lloyds Banking Group’s Consumer Digital Index found that 56% of UK adults, around 28.8 million people, used AI in the past 12 months to help manage their money (reported 3 November 2025).

So a regulated firm has three readers to satisfy at once: Google, the regulator and the AI engines. They want broadly the same thing, which is the useful part.

Why do Google’s YMYL standards weigh so heavily on financial firms?

Because Google says so in its own guidance. For topics that could affect a person’s financial stability, Google states that its systems give even more weight to content showing strong experience, expertise, authoritativeness and trust, and that trust matters most. A page making unverified performance or rate claims gives those systems little to trust, so it tends to lose ground to pages that show their working.

Google’s guidance on creating helpful, reliable, people-first content (last updated 5 October 2026) is careful to say that E-E-A-T is not a single ranking factor. It describes what Google’s systems try to reward. There is no YMYL “penalty” that switches on. The effect is relative: on a money query, your page competes with pages that name their author, date their figures and cite their sources, and Google has said which kind it prefers.

The Search Quality Rater Guidelines, the manual for the people who assess Google’s results, were last revised on 11 September 2025 with clarified YMYL definitions and new examples covering AI Overviews. Rater scores don’t rank pages directly, but they show what Google’s systems are being tuned towards.

What reads as untrustworthy on a money page

  • Performance figures with no reference period, source or date. A raw return figure can’t be checked, so it adds nothing to trust.
  • “From” rates and fees with no “as at” date. Readers and machines can’t tell whether the figure is still current.
  • No named author, reviewer or regulatory details. On a YMYL topic, who stands behind the page is part of the content.
  • Superlatives with nothing to check them against. “Award-winning” or “trusted by thousands” without a source reads as marketing, not evidence.

Each of these is also a problem under the FCA’s rules, which is where the two standards meet.

What do the FCA’s COBS 4 rules mean for search content?

COBS 4.2.1R requires a firm to ensure that a communication or a financial promotion is fair, clear and not misleading. For search, the rule reaches further than many marketing teams expect. It covers web pages, blog posts and social content, and in practice it should also govern the titles, descriptions and structured data that Google and AI engines display on their own, away from your warnings.

Chapter 4 of the FCA’s Conduct of Business Sourcebook (COBS) sets out how authorised firms communicate with clients, including through financial promotions. Three parts matter most for search content. The summaries below are plain-English paraphrases; read the Handbook text before relying on them.

  • COBS 4.2.1R, fair, clear and not misleading. Applies to communications and financial promotions alike (FCA Handbook, COBS 4.2, last updated 23 October 2025).
  • COBS 4.5.2R, a balanced impression. Promotions to retail clients must give a balanced impression of benefits and risks, and must not disguise, diminish or obscure important information or warnings (FCA, risk warnings for mainstream investments).
  • COBS 4.6.2R, past performance. Performance must not be the most prominent feature; it must cover the preceding five years, or the whole period if shorter, in complete 12-month periods; the reference period and source must be clearly stated; and a prominent warning must say that past performance is not a reliable indicator of future results (FCA Handbook, COBS 4.6, last updated 6 April 2026).

Search makes this harder in one specific way. A results page shows a title and a snippet. An AI answer quotes a sentence or two. Neither carries the risk warning sitting three paragraphs further down. A meta title reading “9.4% average annual returns” can therefore appear in public as a performance claim with no period, no source and no warning. The working rule: any sentence that might be quoted alone should be accurate and compliant alone.

The regulator is checking. The FCA’s financial promotions data for 2024 (published 7 February 2025) shows 19,766 promotions amended or withdrawn after it intervened with authorised firms, up 97.5% from 10,008 in 2023. Its 2024/25 annual report puts the 2021 figure at 573. Claims management companies accounted for 9,197 of the 2024 total, so the rise isn’t only about investment marketing, but the direction is clear.

Where do AI answers about financial firms get their information?

Mostly from other websites. McKinsey’s analysis of AI search found that a brand’s own sites make up only 5 to 10 percent of the sources AI search references, and that in financial services more than 65 percent of sources are publishers, user-generated content and affiliate sites. On-site SEO improves the smaller share. The larger share depends on what other sites publish about you.

Put the other way round, roughly 90 to 95% of what an AI engine reads about a typical brand comes from somewhere other than the brand’s own site. McKinsey published the analysis on 16 October 2025, alongside a survey finding that 40 to 55% of consumers in sectors including financial services use AI-based search when making purchasing decisions. It also found that some leading credit card brands were missing from some AI answers in their own category, and that just 16% of brands systematically track their AI search performance.

Clicks are a weak safety net. Pew Research Center found that Google users who saw an AI summary clicked a link inside it in just 1% of visits (Pew Research Center, July 2025). If an answer describes your firm wrongly, or leaves it out, most people will never reach your corrected page.

Shortcuts are riskier than they were. On 15 May 2026 Google added “attempting to manipulate generative AI responses” to the definition of spam in its spam policies, so the rules that govern rankings now cover AI Overviews and AI Mode too. Buying mentions, seeding reviews or mass-producing pages to steer AI answers sits on the wrong side of that line.

What holds up is accuracy at the source. AI engines repeat what they read, so the work is making sure the facts about your firm are correct, current and consistent wherever they appear: your own pages, your FCA Register entry, the comparison and editorial sites in your category, and the directories that describe firms like yours.

Which page habits cost regulated firms visibility, and how do you fix them?

Most visibility problems on financial sites trace back to a handful of page habits that fail Google’s trust standards, the FCA’s rules and AI readability at the same time. The table sets out six common ones: what each looks like to search engines and AI answers, which rule it brushes against, and a fix that satisfies all three readers without adding new compliance risk.

Page habitWhat search and AI seeRule it touchesThe fix
A return figure in the page title or meta descriptionA performance claim shown in results with no period, source or warningCOBS 4.6.2R: prominence, reference period, source and warningKeep titles descriptive. Put figures in the body, with period, source and warning in the same sentence.
Undated rates and fees (“from 4.9%”)AI engines can keep quoting a figure after it has changedCOBS 4.2.1R: a figure that is out of date can misleadAdd “as at [date]” beside every rate and fee, and review on a fixed schedule.
Key figures only in PDFs, images or JavaScript calculatorsCrawlers and AI engines may miss them or read an older copyMainly a readability gap rather than a rule breachPublish core figures and eligibility rules as HTML text or a table. Keep the calculator as an extra.
Benefits in one section, risks in anotherAn AI answer lifts the benefit sentence on its ownCOBS 4.5.2R: a balanced impression of benefits and risksPair each benefit with its main risk or condition in the same paragraph.
Articles with no author, reviewer or regulatory detailsWeak trust signals on a YMYL topicGoogle’s E-E-A-T guidance: trust counts mostName the author and compliance reviewer, show the firm reference number and link to the FCA Register entry.
Bought mentions or bulk AI-written pages aimed at AI answersSpam risk on every Google surface, AI Overviews includedGoogle spam policies (AI manipulation added 15 May 2026)Earn coverage with accurate, checkable facts. Review every page before it goes live.

None of these fixes needs a new marketing claim. Each one makes an existing page more precise, which is usually the easiest kind of change to get through a compliance review.

How much competitive advantage does SEO give a financial services firm?

On the Cllimber Opportunity Index 2026, SEO scores 45.8 out of 100 for financial services firms. That is fourth of six software categories for the sector, behind sales automation (52.3), CRM (49.8) and lead generation (47.3), and slightly below the sector average of 47.5. It sits in the band where the advantage grows the longer a firm keeps at it.

CategoryScoreBandWhere the edge comes from
Sales automation52.350+Wins the long, considered sale
CRM49.843 to 49Client history compounds into repeat business and referrals
Lead generation47.343 to 49A steady flow of qualified enquiries
SEO45.843 to 49Found when a buyer starts researching, then trusted on the page
Marketing45.343 to 49Visibility held across a long buying cycle
Social media44.343 to 49Credibility, if the content is consistent

Source: Cllimber Opportunity Index 2026, financial services breakdown (updated 12 November 2026). Sector average 47.5. Bands: 50+ is a big, lasting edge that rivals struggle to close; 43 to 49 is a strong edge that grows the longer it’s sustained.

The Opportunity Index estimates the competitive advantage a business gains from using a category of tool effectively, compared with a direct competitor that doesn’t. It covers 63 industries and 378 industry-and-category combinations, and it scores categories, not individual products or agencies.

Read for search, the score says two things. SEO isn’t where a financial firm gains its largest edge; the systems that win and keep client relationships score higher. But the edge SEO gives compounds, because trust signals, dated facts and earned coverage build on each other. A firm that starts correcting its pages and its third-party footprint now is building an asset that takes a competitor time to match.

How should a regulated firm build pages that rank and get cited?

Build each page so that any sentence can be quoted alone and still be accurate. In practice that means one topic per page, the answer first, figures in plain HTML with their date and source beside them, benefits paired with risks, named people standing behind the content, and structured data that repeats the visible facts. Then put compliance sign-off around everything search engines display.

  1. One question per page, answer first. Open each section with a direct two- or three-sentence answer. AI engines lift passages rather than whole pages, so the passage needs to make sense alone.

  2. Facts in text, with dates. State current rates, fees, minimums and eligibility rules in HTML text or a table, each marked “as at” a date. Calculators and PDFs can stay, but they shouldn’t be the only place a fact lives.

  3. Qualifiers travel with claims. Put the period, source and warning in the same sentence as any performance figure, and the main condition in the same sentence as any benefit.

  4. Show who is responsible. Name the author, show who reviewed the page for compliance, and give the firm reference number with a link to the FCA Register.

  5. Structured data that matches the page. Mark up the organisation, the article and, where it fits, the product. Schema.org’s FinancialProduct type has properties for interest rate, annual percentage rate and fees. Every value in the markup should match the visible page exactly.

  6. Sign-off covers what search displays. Include page titles, meta descriptions, FAQ answers and schema text in the compliance review. These are the parts most likely to be shown without the rest of the page. Google’s guidance on AI-generated content now makes the same point: since an update on 1 October 2026, its call for a manual fact-check before publishing explicitly covers titles, meta descriptions, structured data and image alt text.

  7. Review on a schedule. Recheck rate, fee and eligibility pages monthly and whenever a product changes, and update the “last reviewed” date when you do.

How can a firm check what AI engines currently say about it?

Ask them the way a customer would, and write down what comes back. Run the ten to twenty questions your prospective clients actually ask through ChatGPT, Perplexity, Copilot and Google’s AI Overviews. Record whether your firm is named, which figures are quoted, which sources are cited and whether anything is out of date. Repeat monthly so that changes show up.

Use questions with buying intent rather than your brand name: “which platforms offer a SIPP with no exit fees”, “how much does ongoing financial advice cost in the UK”, “specialist lenders for contractors”. Answers vary by engine, location and account, so test signed out, note the date and keep screenshots.

A simple log is enough: question, engine, date, firm named or not, figures quoted, sources cited. Patterns appear quickly. If engines quote a fee you changed last year, find the page they cite and correct it there. If they cite a comparison site that lists you wrongly, ask that site to update its entry. Our guide to tracking AI citations sets out a fuller scorecard.

What should financial firms avoid in SEO and AI search?

Avoid anything that trades accuracy for reach. For a regulated firm, the costly mistakes are the ones that put a misleading impression into public view: performance claims in titles, outdated rates left live, bought or manufactured mentions, bulk-generated pages nobody reviewed, and third-party content promoting your products that nobody checked. Each one risks a search problem and a regulatory problem at once.

  • Ranking or citation guarantees. No one controls which sources an AI engine chooses. A supplier promising placement is either guessing or proposing to manipulate results.
  • Bought mentions and link schemes. Since 15 May 2026, Google’s definition of spam covers attempts to manipulate generative AI responses as well as rankings.
  • Unreviewed pages at scale. Google’s spam policies treat pages produced in bulk mainly to rank as scaled content abuse, however they were written.
  • Unchecked third-party promotions. A publisher, affiliate or influencer promoting your regulated products may be communicating a financial promotion. Section 21 of the Financial Services and Markets Act 2000 restricts that unless an authorised person makes or approves it, and since February 2024 firms have needed FCA permission to approve promotions for unauthorised persons.
  • Stale product pages. A page left live after a rate change can mislead readers, and AI engines can keep quoting it.