Category

Aug 3, 2025

AI Visibility for Wealth Managers: The Spear’s 500 Problem

The Spear’s 500 blocks every major AI crawler from its adviser profiles. VouchedFor blocks none. The credential the sector works hardest for is the one ChatGPT cannot read.

Golden hour over a London garden square lined with period townhouses, for an article on AI visibility for wealth managers

Category

Aug 3, 2025

AI Visibility for Wealth Managers: The Spear’s 500 Problem

The Spear’s 500 blocks every major AI crawler from its adviser profiles. VouchedFor blocks none. The credential the sector works hardest for is the one ChatGPT cannot read.

Golden hour over a London garden square lined with period townhouses, for an article on AI visibility for wealth managers

Ask ChatGPT, Gemini, Claude or Perplexity to recommend a wealth manager in London for a client with £5 million, and the AI engine does not consult the industry's own hierarchy. It reads whatever pages it is allowed to read. In UK wealth management, a surprising number of the most authoritative pages are closed to it.

We checked the crawler rules on the three third-party sources that matter most to UK wealth managers and financial advisers: the Spear's 500, Citywire New Model Adviser, and VouchedFor. The results invert the sector's own sense of prestige. The directory listing an adviser works hardest to earn is the one AI models cannot read. The directory many wealth management firms quietly consider downmarket is wide open.

Which sources do AI models actually read?

An AI system answering a question about financial advice has three ways to get information. It can rely on what was in its training data, it can retrieve live pages at the moment of the query, or it can lean on a search index behind something like Google AI Overviews. All three depend on permission, and that permission is declared in a file called robots.txt at the root of every website.

Robots.txt names individual crawlers and tells each one which paths it may read. OpenAI runs GPTBot for training and OAI-SearchBot for live retrieval. Anthropic runs ClaudeBot and Claude-SearchBot. Google separates ordinary search crawling from AI use through Google-Extended. Perplexity runs PerplexityBot. A publisher can allow all of them, block all of them, or do something more selective.

Most wealth management firms have never looked at what the directories they are listed in have chosen. Here is what they chose, as of November 2026.

Why can AI not read your Spear's 500 profile?

The Spear's 500 describes itself as a guide to the best private client advisers, covering wealth managers, family lawyers, tax advisers and property advisers. Its rankings sort advisers into three tiers: Top Flight, Top Recommended and Recommended. Spear's says the rankings draw on peer nominations, client feedback, interviews, firm-supplied data and its own research, scored through a proprietary system. Earning a place takes a submission, referees and a full research cycle.

The robots.txt file at spears500.com blocks GPTBot, ChatGPT-User, Google-Extended, ClaudeBot, PerplexityBot, CCBot, Amazonbot, FacebookBot, Bytespider and Applebot-Extended. It does not block them from the whole site. It blocks them from exactly three directories: /adviser/, /company/ and /school/.

Those three directories are the directory. Every individual adviser profile sits under /adviser/. Every firm profile sits under /company/. The searchable database an adviser submits to be included in is precisely the part no major AI crawler is permitted to read.

So if you are a Spear's-ranked wealth manager, an AI engine asked to recommend good advisers in your field cannot see your profile, your tier, your specialisms or your client base. The ranking is real. It is just not readable.

What about the editorial pages on spearswms.com?

Here the picture improves, and this is the part worth acting on.

Spear's publishes its indices twice. The profiles live on spears500.com, which is blocked. The editorial write-ups live on spearswms.com, the magazine site, which blocks only Meta's crawlers and leaves OpenAI, Anthropic, Google and Perplexity free to read everything.

The Spear's HNW Wealth Managers index page on the magazine site is therefore readable. Which raises the question of what is actually on it.

That page has two parts. The first is a set of prose write-ups of roughly a dozen advisers, each a few paragraphs long, carrying the adviser's name, firm, focus, ranking tier and quotes from an interview. The second is the complete list of all 105 ranked advisers, presented as a sortable table.

The table is an embedded Datawrapper chart served from a separate domain, and that embed carries a noindex instruction. The names inside it are not part of the Spear's page as search engines and AI crawlers index it.

The consequence is stark. Around twelve of the 105 ranked advisers appear in crawlable prose on an AI-readable page. The other ninety-odd appear only inside a noindex iframe, and on a profile page that blocks every major AI crawler. Being ranked is not the same as being visible. Being written up is.

Does Citywire get through?

Citywire New Model Adviser runs the annual Top 100, a long-standing list of UK financial planning firms. Its robots.txt is the most carefully built of the three, and it draws a distinction worth understanding.

Citywire blocks the training crawlers outright across the entire site: GPTBot, ClaudeBot, Google-Extended, CCBot, Bytespider, Amazonbot and others all receive a blanket disallow. It then treats the retrieval crawlers differently. OAI-SearchBot, ChatGPT-User, Claude-SearchBot, Claude-User, PerplexityBot and Perplexity-User are allowed into the /news/ area only, and blocked everywhere else. The file also carries an explicit legal notice stating that use of Citywire content for machine learning or AI training is not permitted.

Because the Top 100 is published as a news article, it sits inside the one area those retrieval crawlers can reach. A Top 100 placement can therefore surface when an AI engine goes and fetches a page in response to a live question. It will not have been absorbed during training, because the training crawlers never saw it.

That distinction matters more than it sounds. Training-time exposure is what makes a model name your firm unprompted. Retrieval-time access only helps when the engine decides to go and look, and when your page is among the handful it fetches.

Why is VouchedFor the one that gets through?

VouchedFor is a review platform for financial advisers, mortgage advisers, solicitors and accountants. It verifies reviewer identity, checks that each adviser is authorised by the appropriate governing body, confirms key details such as fees and qualifications, and says it never removes or edits negative reviews. At the high-net-worth end of the market it tends to be read as a mass-market consumer site rather than a professional credential.

Its robots.txt allows everything. There is no AI crawler block at all. It publishes four sitemaps, one of them dedicated to adviser profiles, which is an open invitation to crawl every profile on the platform.

It also lets advisers publish articles under their own name on their profile, which puts an adviser's words, on an adviser's named page, on a domain every AI crawler is free to read.

The sector's ranking of prestige and the ranking of AI readability are close to inverted.

Does being in the Spear's index still matter?

Yes, and the point is not to stop pursuing it.

Spear's placement matters for the reasons it always has. Referral sources read it, journalists use it as a shortlist, and clients at the top of the market recognise it. None of that depends on a crawler.

What has changed is that it no longer does the second job firms assume it does. A directory listing used to be a durable, authoritative, third-party mention on a high-authority domain, which is exactly the kind of signal that shapes an AI model's sense of who is credible in a field. Spear's has chosen, deliberately, not to let that happen on its profile pages, and many publishers have made the same call for sound commercial reasons: their content is their product.

The firms that lose out are the ones treating the listing as though it still carried that second benefit, and doing nothing else.

What should a wealth management firm do about it?

Aim for the write-up, not just the ranking. The gap between appearing in prose on the magazine page and appearing only in the index table is the gap between visible and invisible. Spear's conducts interviews as part of its research and quotes advisers in the index write-ups. Being available, quotable and specific during a research round is what gets you into the readable part of the page.

Submit to the research rounds, and submit early. Submission is free and open to every private client firm. Spear's publishes a full research calendar with dates for each index. In the 2026 cycle, the HNW Wealth Managers, UHNW Wealth Managers and Family Office Services indices opened for research in February, closed to submissions in mid-March, and launched at the end of May. If your firm wants to be in the next round, assembling the submission and lining up client referees belongs in the diary now, not in February.

Treat the open directories as infrastructure, not as a downgrade. A verified, complete, well-reviewed profile on a platform that permits AI crawling is a cheap and durable visibility asset. It will be read. The prestigious one will not.

Announce the placement on your own site. When a ranking lands, publish a page on your own domain naming the adviser, the index, the tier and the year in plain text. Your own website is the one source whose crawler rules you control. If the directory will not let an AI system read that you are Top Flight, your own site can.

Audit your own robots.txt before anything else. We regularly see firms spend a year chasing third-party credibility while their own website quietly blocks the same crawlers, usually through a default inherited from a web platform or a security product nobody chose. It costs nothing to audit and it is the first thing to fix.

Publish the facts an AI engine needs in order to recommend you. Who the advisers are, what each one specialises in, the minimum portfolio the firm works with, where it is based, the fee structure, and how a prospective client starts. Put it in readable text with structured data behind it, rather than in a PDF or a brochure download. An AI system recommending a wealth manager needs the same things a prospective client needs.

Keep the Google Business Profile accurate. Local results still feed both traditional search and AI answers for location-shaped queries, and a stale address or an old phone number on the profile is one of the most common reasons an AI model describes a firm wrongly.

Frequently asked questions

Will ChatGPT recommend my wealth management firm? Only if it can read enough about you from sources it is permitted to crawl. For most UK firms the binding constraint is not quality of advice but availability of text: an FCA-authorised firm with a thin website and a blocked directory profile gives the model almost nothing to work with.

Is this SEO or GEO? The work overlaps but the target differs. Traditional SEO aims at a ranking position. GEO, or generative engine optimisation, aims at being the name an AI engine produces when there is no list of ten blue links. Crawler permissions, structured data and answer-shaped content serve both.

Does the FCA Register help? It confirms you exist and are authorised, which matters for verification, but it says nothing about what you specialise in or who you suit. It is a floor, not a differentiator.

How often should this be checked? Quarterly is reasonable. Robots.txt files are edited without announcement, and a source that is readable this quarter can be closed the next.

The wider point

Every sector has gatekeeper sources that AI models lean on, and in almost every sector those sources have made their own quiet decisions about who may read them. Those decisions are published, they are free to check, and almost nobody checks them.

Before you spend another year pursuing a credential, it is worth thirty seconds confirming that the place it will be published is a place an AI engine is permitted to go.

Sources and method

Crawler permissions were read directly from the robots.txt files at spears500.com, spearswms.com, citywire.com and vouchedfor.co.uk in November 2026. Index methodology, tier definitions, submission process and research calendar dates come from the Spear's research guidelines and the Spear's Wealth Management Indices page. Vetting and review policy comes from VouchedFor's own description of its process. Robots.txt files change without notice, and the rules described here may have changed since publication, which is precisely why checking them belongs on a recurring schedule rather than in a one-off audit.

More Insights

Read Next

Find out what AI tells your clients about you.

Find out what AI tells your clients about you.

Find out what AI tells your clients about you.

Subscribe to the Briefing

Your information is kept confidential

Recommendable

Subscribe to the Briefing

Your information is kept confidential

Recommendable

Subscribe to the Briefing

Your information is kept confidential

Recommendable