Technical and on-page SEO for law firm sites

The three failures that actually cost law firms rankings

After a while, even a badly broken site starts to look familiar in its own boring way.

1. A redesign with no redirects. Google cannot follow the old URLs and you are back to square one. So you end up with a new and better-looking site and nobody notices all the damage that's behind it.
2. An indexing directive in the wrong place. Your staging site shouldn't be accidentally ending up within Google's index. Or the live template is simply not indexable after launch. Sometimes the single noindex line renders entire portions of a site invisible.
3. A migration that changes the practice area URL pattern. So many agencies forget to update the internal links. Of course you don't know it's a practice page until you reach the end of the three-hop redirect chain.

These failures stay there broken for months and nobody notices them then. The traffic drop comes first, the agency's algorithm excuse second. Usually it's the redirect map that really explains the sudden drop.

Core Web Vitals, as they stand now

There are three main metrics in Google's Core Web Vitals¹:

Interaction to Next Paint grew to replace the old First Input Delay metric in 2024 and Google never looked back¹.

Don't worry too much about your speed scores or mobile test results. The Hennessey Digital study covers nearly 16,000 law firm landing pages in 200 city markets. It seems clear to them there are both good and poor mobile speed scores among high-ranking legal sites and that sometimes the popularity may outweigh the low score⁴.

WordPress, themes, and the replatform pitch

From the same study⁴:

• 65% of law firm sites run WordPress
• Almost all of the ranking ones use an SEO plugin
• No correlation between the CMS and average position

Choosing WordPress is not the problem. Although a new platform is probably pitched as a ranking factor, you can't conclude that from this study.

What makes these sites particularly fragile is the steady build-up of plugins and settings:

• Three SEO plugins with overlapping title rules
• A page builder rendering headings as styled divs
• A caching layer serving a stale canonical
• An old sitemap plugin publishing URLs that were deleted two years ago

We preferred a one-day conflict cleanup than a month-long migration with a new platform.

What schema can honestly do for a law firm

Google's search gallery is the full list of structured data features, published by Search Central. None of these features have anything for lawyers, legal or professional services. There is no legal rich result.

Local business schema is officially supported and the markup can result in hours, ratings, directions and booking actions in the knowledge panel⁵.

Site structure, URLs and internal links

The CMS may publish your practice pages, but I need them moved around into certain folders in the firm's own structure. Target a single, real client search per page (relevant to your practice). Having a city page is necessary only if you are saying something real about that city.

I've seen a few state-wide firms build city pages to rank for specific, similar queries and Google pretty much treats them as doorway abuse¹⁰. The content page has the full city page test.

Internal linking is cheap to fix, and it is the most neglected job on a legal site:

• A link from every blog post that touches a practice area to that practice page, with the anchor a human would write
• Links from the attorney page to the practice pages they handle, and back

I click through larger firm sites and often notice the converting pages four clicks from the homepage. They are linked only from the navigation menu.

What AI retrieval needs from your markup

There is no special requirement here and pages routinely appear in AI features without adding any special files or markup. The one thing that is documented by Google is an indexed page that is eligible to be shown with a snippet¹¹.

It's worth checking what four settings will do:

• A nosnippet directive, usually added years ago by somebody worried about scraping
• A restrictive max-snippet value
• data-nosnippet wrapped around the opening paragraphs, which is the part that gets extracted
• AI crawlers blocked in robots.txt or at the firewall, which some legal hosts and security plugins do by default

Oftentimes the firm is quietly removed from AI answers once the wrong setting reaches the live site.

It's true, AI retrieval is pretty literal.

• Dan Petrovic's work on grounding snippets, with exact sentences lifted rather than paraphrased and opening paragraphs extracted close to wholesale¹²
• AirOps has found that ChatGPT discards 85% of the pages it retrieves, with the cited share right around 15%¹³

That means "boring" markup and "old-fashioned" rules, not clever rendering or script tricks.

1. Text in the HTML, not assembled by script after load
2. Real headings, not styled divs, because the heading is the matching surface
3. Content behind tabs and accordions still present in the markup

For a page that needs JavaScript to show its first paragraph, its most valuable sentences are remarkably out of reach. The writing side has its own page: writing for AI page.

The redesign checklist, including the line nobody has

1. Use your crawl and Analytics data to spot every URL with an entrance or a conversion in the last twelve months to find the pages that are worth redirecting.
2. I use the export to check for gaps, but do all of the URL mapping by hand. Each unmapped URL is a decision, not a homepage redirect.
3. New practice area URL patterns are only useful if they are genuinely needed. A tidier structure is expensive in reset rankings.
4. Block the site properly on staging, check it, then unblock it on production.
5. Compare the titles and copy page by page, heading by heading, and you'll find the gaps. Redesigns are for looks, and the text they lose is very specific to the firm.
6. Make sure you have the required disclaimers and the responsible lawyer and office details on the new site. This is the rarest line on a redesign checklist you'll ever see.
7. Re-submit the sitemap. Then watch the daily coverage figures in Search Console come rolling in for a fortnight.

What I check first, in order

1. Indexation against the page list. No point in looking for problems in pages Google has not indexed.
2. Redirects, for chains and loops, then canonicals against the URLs that actually rank.
3. The four snippet and crawler settings above. They generally are binary and free to repair.
4. Titles and heading structure on the pages that matter, and whether the headings are really headings.
5. Mobile speed. Don't spend time on the whole site, look at the top five entry pages.
6. Schema validity, and whether it says anything true.
7. Internal links into the practice pages, counted rather than assumed.

A full check usually takes me a day on a normal firm site, and we're three days in on one that has been through two agencies.