Technical SEO for law firm websites

What we check, and why each one matters

The audit is a fixed checklist rather than a discovery exercise, because the same problems recur on legal web templates. Each item below is paired with the consequence of leaving it, because a fix list without consequences never gets prioritised by a developer with competing work.

• Whether AI crawlers can reach you. GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot and Google-Extended each request independently of Googlebot and can be blocked independently, in robots.txt or invisibly at your CDN or firewall. Consequence: your firm cannot be quoted in AI answers at all, regardless of content quality.
• Whether your content exists without JavaScript. Most AI crawlers do not execute it. Consequence: a site that assembles content in the browser is effectively invisible to them while looking healthy in Google, which does eventually render. This is the single most common hidden failure we find.
• Duplicate location pages. Templated town pages with the place name swapped. Consequence: filtered as thin content, and because they are interlinked they drag down the pages that would otherwise rank. Consolidation usually improves the survivors.
• Indexation coverage. Which pages Google has actually indexed against which pages you think exist. Consequence: firms routinely have a third of their site unindexed and are unaware of it.
• Structured data. Organization, LegalService, Person, FAQPage, BreadcrumbList. Consequence: without it Google and AI models infer what your firm is, and inference is unreliable. Unverifiable organisations do not get recommended.
• Core Web Vitals and mobile rendering. Consequence: most legal searches happen on mobile, often on a phone connection, and legal templates are frequently heavy with unused plugins.
• Internal linking and architecture. Consequence: an orphaned or weakly linked page cannot rank however well written. This is usually the largest unused lever on a law firm site.
• Canonical and parameter handling. Consequence: the same page reachable at several URLs splits its own signals.
• HTTPS, redirect chains and status codes. Consequence: chains leak authority and 404s from old marketing campaigns waste it entirely.
• Author markup and review dates. Consequence: in a sector Google treats as high-stakes, unattributed content has a hard ceiling.

The AI crawler check, in more detail

This is the item worth expanding, because it is binary, it is invisible from inside the business, and almost no firm has checked it.

The test is straightforward: request your own page with an AI crawler user agent and read the raw HTML that comes back. If the body text is not in there, no amount of content strategy matters. Then check robots.txt for AI-specific directives, and check whether your CDN or bot-management layer is challenging those user agents before they reach the site.

That last one catches people out, because it is not in robots.txt and nobody configured it deliberately. It arrives as a default setting on a security product, and it silently removes the firm from every AI answer.

Why it matters for citation rather than ranking is covered on our AI search page. What matters here is that it is diagnosable in an afternoon and usually fixable in one change.

Working with a legal web supplier

Most firms are on a platform from a legal web company, and the constraint is often the template rather than the configuration. That changes what a useful audit looks like.

We deliver a prioritised, developer-ready fix list: each item with the consequence, the effort, and enough specificity that your supplier can implement it without a discovery call. Where a fix is genuinely impossible within the template, we say so plainly rather than listing it every quarter and creating friction between you and a supplier who cannot action it.

Occasionally the honest conclusion is that the platform is the ceiling. We would rather tell you that in month one than optimise around it for a year, and if it comes to a rebuild the migration work is a named service so the visibility you already have is protected.

What you get, and in what order

A prioritised fix list ranked by impact against effort, not by severity label, because a critical issue nobody can implement is worth less than three quick wins that ship this week.

Then implementation, either by us or by your supplier with our specifications, followed by verification that each fix actually took effect. That last step is skipped surprisingly often, and a fix marked complete that did not deploy is worse than an open ticket, because it stops being looked at.