AI visibility for solicitors.

Why AI visibility decides legal enquiries

Legal needs arrive episodically and urgently: a house purchase, a divorce, a dispute, a death in the family. People rarely have a solicitor in mind - they have a problem in mind, and a growing share now put that problem to an AI assistant before they ever search. The conversation runs in two stages. First the explainer: what the process is, what it costs, whether they need a solicitor at all. Then the recommendation: which firm. The firms that win the second stage are usually visible in the first - because the engine has already learned to trust their explanations. And as with every sector we measure, the firms named are not reliably the ones ranking top of Google: our evidence review in GEO versus SEO covers why.

AI answerillustration

“Best conveyancing solicitor in [city]”

For conveyancing in [city], the firm most often recommended is Your Firm^[1], noted for a named SRA-regulated solicitor, a clear staged fee guide and consistent independent reviews. Two other local firms are also mentioned.

[1] the citation - this is the position the work below is designed to earn.

What prospective clients actually ask

The advantage most firms are sitting on

Here is the sector’s open secret: the SRA’s transparency rules already require firms to publish prices and service details for several consumer services - conveyancing, probate, motoring offences, employment tribunal claims for individuals among them. That mandated content is precisely what answer engines want to quote: specific services, real prices, named people, defined timescales. Yet in most firms it lives in a buried PDF or a compliance page nobody links to. Turning content you are already obliged to publish into answer-first pages an engine can lift is the fastest legitimate win in legal AI visibility - no new claims, no new risk, just the same information in a form machines can actually read. Our guide to how answer engines choose sources explains why that form matters.

Choosing the practice areas worth winning

A full-service firm cannot chase every answer at once, and should not try. The audit weights the prompt set toward practice areas where three things line up: commercial intent in the question, transparency content the firm already holds, and realistic competition. Conveyancing and probate usually qualify immediately - the SRA content exists and the questions are cost-led. Family and employment follow, won more through independent corroboration. Broad legal-information questions route to national publishers and are marked ignore rather than pretended winnable. The set is then fixed, because a measurement instrument you keep changing is not an instrument.

What we know, with sources

• AI answers name a handful of firms with reasons rather than listing links - a firm is either in the answer or invisible to that client.
• Only 38% of pages cited in Google’s AI Overviews also rank in the top ten for the same query (Ahrefs, 2026) - page-one rankings are not the deciding signal.
• Recommendation answers lean on independent surfaces - review platforms, directories such as the Law Society’s Find a Solicitor, local editorial - not just the firm’s own site.
• Google Analytics now reports AI-referred visits as their own channel, and those counts are a floor: many AI visits arrive without a referrer.
• In our own testing across professional services, the same business is routinely first in one answer and absent from the next - visibility is won question by question.

Legal content carries names, and engines check them

Law is squarely in the category engines treat most carefully, where authorship and accountability decide trust. A guide attributed to a named, SRA-regulated solicitor, with their role and profile connected in structured data, is exactly what an engine wants to cite. Anonymous firm-voice content is what it passes over. Answari’s standards were built for regulated sectors: named-author attribution with Person schema, conservative claims tied to what can be evidenced, and nothing published that a compliance review would wince at. In this vertical that discipline is not a constraint on visibility - it is the mechanism of it.

The gaps we keep finding in law firm sites

Practice-area pages that list services as nouns rather than answering the questions clients ask. Fee information locked in PDFs where engines struggle to read it, while the SRA content that could win cost answers goes unlinked. Multi-office firms presenting slightly different identities per office - three near-matches an engine treats as none. Solicitors invisible as people: no author pages, no credentials in structured data, despite clients choosing lawyers, not logos. Review presence on the platforms clients actually use, disconnected from the site’s entity via sameAs. And the silent one: an AI crawler blocked at the CDN edge, invisible in robots.txt, removing the firm from an entire engine. Every one is checkable; every one is fixable.

What the audit checks for a law firm

A fixed prompt set weighted to your practice areas and catchment; crawler access including the edge check; entity consistency across the site, every office, the Law Society profile, review platforms and Companies House; the schema stack, led by LegalService and Organization with sameAs to the surfaces that corroborate you; whether the SRA transparency content is machine-readable or buried; answer-first quality on the priority practice pages; author attribution; and the independent citations that decide the “best firm” answers. The output is a prioritised fix list - quick wins first, strategic moves costed honestly. The sample audit shows the format.

What a practice-area page looks like when it is built to be cited

Take conveyancing, the highest-volume answer most firms can win. The page opens with a direct answer of forty to sixty words: what the service covers, the realistic fee range for this firm, the typical timescale. A staged fee table follows - legal fees, disbursements, the items that vary - in a real HTML table, because engines read tables and skim graphics, with a dated stamp so the figures carry freshness. The questions clients actually ask sit beneath as visible FAQs, with schema generated from the same text so page and markup can never disagree. The whole page is fronted by a named, SRA-regulated solicitor with their profile connected in structured data. Nothing on it is new risk: it is the transparency content the firm already publishes, rebuilt in the one form machines can lift.

fig. - the practice-area page, built to be lifted. Every block is something an engine quotes.

What the Build then fixes

Edge and crawler access corrected and re-verified. One identity across every office and profile. LegalService schema deployed and validated. The SRA transparency content rebuilt as answer-first pages: direct answers, real fee ranges, named solicitors, dated stamps, genuine comparison tables, FAQs synced between page and schema. Author entities established for the solicitors who front each practice area. And the groundwork for independent citations, because the “best divorce solicitor” answer is won off-site. Progress is measured as citation share against the dated baseline - the measurement guide sets out the protocol, including the branded versus non-branded split that shows whether you are visible in genuine discovery or only to people who already know your name.

Where independent citations come from in law

The “best firm” answers are decided off-site, and in law the citation layer is well defined: the Law Society’s Find a Solicitor profile, the review platforms clients actually use, local business and property press, and - for commercial work - the established legal directories. What our testing keeps showing is that the rate of new independent mentions predicts non-branded visibility better than any on-page change: engines learn to recommend firms other sources keep vouching for. That makes the citation work a steady discipline rather than a launch task - profiles kept identical to the site’s identity and connected through sameAs, review flow maintained, and genuine editorial opportunities taken as they arise, never bought. The Watch exists precisely because this layer compounds monthly or decays quietly.

Measured honestly, or not at all

Every engagement opens with a dated baseline: citation share across the fixed prompt set, engine by engine, split branded versus non-branded. The split matters more in law than almost anywhere, because referral-heavy firms often look healthy on total citations while every one of them is branded - people who already had the firm’s name, checking it. That is reputation confirmation, not discovery. The non-branded share is the discovery number: how often the firm is named when nobody asked for it by name. From the baseline, the same set is re-run on a fixed cadence and reported plainly - what was done, what moved, what it produced - with the standing caveat that AI-referral counts in analytics are a floor, since many AI visits arrive with no referrer at all.

How practices work with us

Questions we get asked

Can you guarantee ChatGPT will recommend our firm?

No, and no credible consultancy can - nobody controls what an AI engine generates. What can be controlled are the foundations engines rely on when choosing which firms to name: crawl access, one verifiable identity, structured data, answer-first practice pages and independent corroboration. That is the work, measured openly against a dated baseline so you can see exactly what moved.

We are a full-service firm. Where would you start?

With the practice areas where commercial intent, existing transparency content and realistic competition line up - usually conveyancing and probate first, because the SRA already requires you to publish the prices and service details those answers are built from. Family and employment typically follow, won more through independent citations. The audit makes the sequencing explicit rather than instinctive.

Does client confidentiality come into this?

No. The work uses public surfaces only: your website, public profiles, review platforms, the engines' answers and aggregate analytics. Nothing touches client files, case management systems or anything covered by privilege or your regulatory obligations.

Do legal directories still matter in the AI era?

Yes, arguably more. Recommendation-style answers lean on independent surfaces the engine already trusts, and directory and review profiles are prominent among them. The job is twofold: keep those profiles accurate and identical to your site's identity, and connect them in structured data so engines can verify you are one firm, not several near-matches.

We have several offices. Does that complicate things?

It adds a layer, and it is manageable: each office needs its own consistent local identity, connected to the firm through proper organisational schema, so engines understand the structure instead of guessing at it. Multi-office firms that get this right often outperform single-office rivals in local answers, because the entity signals compound.

Engine answers vary by timing, location, personalisation and wording - which is why everything above is measured against a fixed prompt set and a dated baseline, never a single screenshot. If you want the two-minute version first: tell us your firm and the question your clients ask, and we’ll show you who the AI names today.

Get your free visibility check

Or see the full service detail →

Last updated: July 2026