We Tested 110 Law Firms For AI Discoverability. 50% Failed.

We analysed 110 UK law firm websites for AI discoverability. Only 50% score above 60 out of 100, and one in six firms have sites that AI search engines struggle to read or recommend. Here's why and what it could be costing you.

Here's the scenario: a prospective client needs an employment solicitor after a tribunal claim lands. They open ChatGPT and type: "best employment law firms in Leeds."

Your firm has handled 400 tribunal cases in the past decade. You are the obvious choice.

You don't appear. Your competitor, two years old, half your headcount, gets recommended. The prospect books a call with them that afternoon.

You never knew they were looking.

We analysed 110 UK law firm websites for AI discoverability in March 2026. Each site was assessed using AireStream's AI Discoverability Score methodology, a structured analysis of the signals AI search engines use when deciding which businesses to recommend.

The results of our research expose a structural gap running through the UK legal sector. 50% of firms score below 60 out of 100. 15% are effectively invisible, scoring under 30. The firms that do appear are accumulating enquiries the rest will never see.

Here is what we found, which parts of the profession are most affected, and why the gap is widening.

The Data: What 110 Law Firms Look Like to AI Search Engines

Overall Discoverability Breakdown

Half of UK law firms we tested fail to reach the 60-point threshold for consistent AI recommendation. The distribution tells a clear story about a sector that has not yet adapted to AI-driven buyer research.

The breakdown across all 110 firms tested:

• High discoverability (60-100): 50% of firms. These firms have the structural signals AI recommendation engines look for: specific content, third-party citations, and clear entity data.
• Medium discoverability (30-59): 35% of firms. These firms have partial signals in place but gaps across schema, content depth, or third-party presence that make recommendations inconsistent.
• Low discoverability/Invisible (0-29): 15% of firms. These firms do not appear in AI-generated shortlists. Some sites are completely inaccessible to AI crawlers.

One in six UK law firms are invisible when a prospect researches legal services using AI. That is not a technology problem. It is a positioning and content problem with a commercial price.

The spread within the high-discoverability tier is also instructive. Only two firms scored above 80. The majority of "high" scorers cluster between 60 and 75, discoverable, but with meaningful room to move up and dominate AI recommendations in their practice area.

Where the Discoverability Gap Is Deepest: Sub-Score Analysis

The overall score is built from four components. Understanding which components fail reveals exactly where the sector is falling short.

Entity Clarity (average 81% of maximum score): Most firms perform well here. AI tools can identify what a firm does, where it operates, and which practice areas it covers. This is the baseline, and most firms clear it.

Technical Foundations (average 98% of maximum score): Websites are technically accessible. Crawlability is not the issue.

Content Structure (average 66% of maximum score): A more significant gap. Missing alt text on images, thin service page copy, and poorly structured case studies reduce AI systems' ability to extract service-specific expertise.

Schema Markup (average 36% of maximum score): This is the defining weakness. Twenty-seven per cent of reachable firms have zero schema markup. Fifty-four per cent score below half marks. Schema is the structured data layer AI platforms use to verify business details, services, and location. Without it, even firms with strong content are harder for AI to confirm and recommend.

The pattern is consistent: firms know who they are and their sites are technically sound. The gap is in how they communicate their expertise in machine-readable formats that AI systems trust.

Score Range and the Reachability Problem

Ten of the 110 firms tested scored zero. In every case, the cause was a site that could not be reached: domains that had lapsed, sites blocking automated requests, or pages timing out.

An AI tool attempting to recommend a law firm with a non-responsive website will simply not include it. The firm is invisible not because of content quality, but because the infrastructure has failed.

This is a different problem from low schema scores or thin content, but it is just as commercially damaging.

Among the 100 reachable firms, the average score was 57 out of 100. The highest score was 81. The lowest reachable score was 8. That 73-point spread between best and worst, among firms all operating live websites, shows how differently the sector is positioned for AI discovery.

Why Are 50% of Law Firms Failing the AI Discoverability Test?

Three structural gaps explain most of the underperformance. They are not technical mysteries. They are decisions, or more precisely, decisions that have not yet been made.

Missing structured data

Schema markup is the layer of machine-readable code that tells AI platforms what a business is, where it operates, and what it does. For law firms, the relevant schemas include LegalService, LocalBusiness, and Organization.

When a prospect asks ChatGPT "employment solicitors Manchester," the AI draws on sources it can verify. Schema markup provides that verification layer. Without it, even accurate website content is harder for AI to extract and trust.

Twenty-seven per cent of the firms we tested have no schema markup at all. Another 27% have partial implementation. Combined, more than half the sector is operating without the structured data layer that AI recommendation engines depend on.

Expertise that looks generic

Legal websites frequently describe services in the broadest possible terms. "We handle employment matters" is useful to a human reader who can ask follow-up questions. It is not useful to an AI tool trying to match a firm to a specific query.

Firms appearing consistently in AI answers have content that signals specific expertise: tribunal types handled, sector clients served, outcome examples. "We defended a manufacturing employer in a £1.4M unfair dismissal claim" gives AI something concrete to index.

Generic service descriptions are a legacy of web content written for Google keyword matching. AI tools need different signals. Specificity, context, and case-level detail are the content factors that convert into AI recommendations.

Absent third-party validation

AI platforms are cautious about recommending businesses they cannot independently verify. For law firms, the verification signals come from external sources: Legal 500 profiles, Chambers rankings, Law Society accreditation pages, court records, and industry publications.

Firms with strong entity clarity scores but weak overall scores typically have the same profile: clear on their own site, invisible everywhere else. AI tools treat third-party citations as trust signals. If a firm's expertise is confirmed only on its own website, the AI has limited grounds to recommend it to a prospect researching options.

Building an external citation footprint takes time. Firms that start now will compound that advantage over the next 18 months as AI search adoption continues to increase.

What Invisibility Costs You

A single employment matter could generate £15,000-25,000 in fees. A corporate transaction or commercial dispute runs considerably higher. These are not marginal numbers.

Firms appearing in AI answers report that 30-40% of new enquiries now come from prospects who researched via ChatGPT or Perplexity before contacting them. These prospects arrive warm: they have already formed a view about which firms are credible before picking up the phone.

The invisible firms are not losing those enquiries to worse firms. They are losing them to better-positioned firms. The prospect never knew the invisible firm existed.

The acquisition cost comparison makes this concrete. For firms that appear in AI answers, these are zero-cost inbound leads. For invisible firms, reaching the same prospects requires paid search, directory listings, referral networks, and business development time.

Your competitors scoring in the top half of this dataset are building a pipeline you cannot see. The enquiries do not show up as lost: they simply do not arrive. By the time the pattern becomes visible in your revenue data, those prospects have already signed with someone else.

What the Discoverable Firms Do Differently

The top-scoring firms in this dataset share three consistent patterns. None of them require significant budget. They require attention and a shift in how content is treated.

They write for AI, not just for humans

High-discoverability firms publish content that is specific enough for an AI system to extract a credible recommendation from. Case studies include outcome details, fee ranges, and client sector. Service pages describe the types of work handled, not just the general practice area.

The firms scoring above 70 in our dataset consistently had service pages with more than 400 words of specific, structured content, not boilerplate. AI platforms index depth of expertise. Generic pages do not demonstrate it.

They maintain an external presence

Discoverable firms keep their Legal 500 and Chambers profiles updated. They contribute to Law Gazette and profession-specific publications. Their partners are quoted in trade press on relevant topics.

These external mentions function as verification signals for AI tools. When multiple independent sources confirm a firm handles specific work in a specific location, the AI has grounds to recommend it confidently. Firms relying solely on their own website for discoverability are operating on one source of truth.

They treat schema markup as infrastructure

The highest-scoring firms have implemented schema markup across their site, not as an afterthought, but as a foundational part of how their site communicates with machines. LegalService schema, LocalBusiness schema, and structured contact data are all present.

This is a one-time technical investment that compounds over time. Firms that implement it now will be indexed more reliably as AI search adoption grows. Firms that defer it will find the gap between their real capabilities and their AI-visible capabilities continues to widen.

Research Methodology

We analysed 110 UK law firm websites in March 2026 using AireStream's AI Discoverability Score methodology. Each site was assessed against the signals that determine whether a business gets recommended by AI search engines including ChatGPT, Perplexity, Claude, and Gemini.

Assessments covered:

• Entity clarity: how clearly the firm communicates identity, location, and services to AI systems.
• Schema markup: presence and completeness of structured data.
• Content structure: depth, specificity, and AI-readability of service content.
• Technical foundations: site accessibility, crawlability, and metadata.

Firms were identified from UK legal directories and professional networks spanning multiple practice areas. Ten firms scored zero due to unreachable domains at time of testing. These are included in the overall count but flagged as infrastructure failures rather than content failures.

Scores represent a snapshot of AI discoverability at the time of testing in March 2026.

Check Where You Stand

If your firm is in the 50% scoring below 60, the gap to the discoverable firms widens every week prospects research without finding you.

Our AI Discoverability Score tool analyses your website against the same criteria used in this study. Enter your website URL and email to get your instant score and the priority fixes that will move your number.

You will see:

• Your discoverability score (0-100)
• Which AI discoverability signals you have in place (and which are missing)
• Your specific gaps versus high-scoring competitors
• One priority fix to improve discoverability immediately

For a detailed picture of your firm's AI discoverability gaps and a roadmap to close them, book a free Growth Assessment.

We test over 100 high-intent prompts across ChatGPT, Gemini, Claude and Perplexity with questions such as "best law firm in [your city]" and "I need legal advice, which law firm in [your city] is most reliable" to show you exactly where you are invisible, why, and what the firms being recommended regularly in your practice area are doing differently.

The firms that address this now will own AI discovery in their market. The ones that wait will spend the next two years wondering why inbound volume dropped.