Law firm SEO: practice area pages, local signals, and content depth

Practice area pages don’t rank because of word count. They rank because the site’s architecture routes internal authority into them, and most firms are told the opposite. The advice column debate over 500 versus 1,200 versus 2,000 words is arguing about the paint job on a car with no engine.

This piece is for attorneys and marketing leads at small-to-midsize firms who have been quoted a retainer and want to understand what the money actually buys. The model that holds position one for competitive legal queries is a three-tier one: pillar practice area, sub-practice case type, and city page, each earning its keep as a separate document.

Does SEO actually work for law firms, and when it doesn’t

Yes, SEO works for law firms. It also fails four specific ways, and the field ducks all of them.

First, a saturated market with a sub-$3,000/month budget. In a metro PI or family law SERP, that retainer covers audit work and not much else; the firm would be better served running Local Services Ads until the budget can carry an eighteen-month engagement.

Second, a runway shorter than the timeline. The competitor consensus is 60 to 120 days to measurable movement and four to eight months to practice-area rankings. A firm expecting signed cases at day 90 is going to fire the SEO at day 91.

Third, a practice area with genuinely low search demand. Some niches convert on referral and paid, not organic, and no amount of content depth changes that. Check the SEMrush volume before the retainer starts.

Fourth, no local presence to legitimately claim. City pages for jurisdictions the firm doesn’t actually practice in get flagged as doorway pages, and rightly so.

The 20% of pages and practice areas that drive 80% of signed cases

The 80/20 map isn’t a rule of thumb. It’s a spreadsheet, and the columns are boring.

• Practice area (the pillar)
• Market (the city or metro)
• Monthly search volume from SEMrush
• Current impressions and average position from Google Search Console
• Competitor SERP structure (what’s ranking and why)
• Estimated case value in that practice area
• Implementation effort (dev-heavy, content-heavy, or both)

Run the rows for every practice area and every market the firm serves, and the pattern is almost always the same: two or three practice-area/market pairs are doing the work, another handful support them, and roughly half the URLs on the site either shouldn’t exist or should be quietly rolled into the pages that do.

Most legal SEO audit deliverables we’ve reviewed hide the two or three fixes that would actually move signed cases inside forty-item lists. We’d rather ship a five-row spreadsheet that names what to fix, in what order, against what expected outcome. The rest of the site exists to route authority to the money pages, not to compete with them for the same query.

Practice area architecture: three tiers, three separate crawl problems

The load-bearing model is pillar practice area, sub-practice case type, and city page. Three separate documents, each with a canonical target of itself. Get this right and the rest of the SEO is content depth and patience. Get it wrong and no amount of copywriting fixes it.

The mechanic is the same one we solved for a national outdoor gear brand: product listing page, then sub-collection, then localized page. Same crawl trap, different vocabulary. When the sub-collection layer doesn’t exist, everything piles onto the pillar and nothing ranks.

The pillar practice area page’s job is to define the practice, not to rank for the city

The pillar page (e.g., /family-law/) is the practice area at the conceptual level. It covers what the firm handles, links to every sub-practice and every genuinely-served city page, and carries named attorney authorship. It should not be trying to rank for “family law attorney Atlanta.” That’s the city page’s job.

Sub-practice pages are where case-type queries actually convert

Sub-practices (/family-law/divorce/, /family-law/custody-modification/, /family-law/prenuptial-agreements/) are where the searcher’s actual question gets answered. Statutes cited with jurisdiction, process walked in order, timeline given honestly, fee structure named. This is where signed cases come from because this is where the searcher was actually looking.

A city page has to name the courthouse or it isn’t a city page

Three-quarters of the “city pages” we audit in legal are city-swap duplicates the crawler treats as one document. Same content, city name changed in three spots. If the firm can’t name the local courthouse, the judges, a jurisdictional filing quirk, or verdict data specific to that county, we’d rather delete the page than rewrite it. Google’s canonicalization will pick one and drop the rest anyway.

Internal linking is the mechanic behind “link to attorney bios”

Every ranking piece in the legal SEO SERP says “link to your practice area pages from your blog” and stops there. That’s not internal linking. That’s a suggestion.

Real internal linking at scale is a spreadsheet. Priority keywords in one column, target URLs in another, anchor text variants in a third. On a national gear brand we worked with, our developer built custom logic that imported that CSV and wired dynamic contextual links from relevant product pages back to their category pages using the priority-keyword-to-URL mapping. Over the four-year engagement, non-branded impressions on those templates roughly doubled, and the highest-value queries moved to the top of page one and stayed there.

The same mechanic applies to a law firm. Blog posts on “how child support is calculated in Georgia” link back to the divorce sub-practice on the phrase “Georgia divorce”; the divorce sub-practice links back to the family law pillar; the pillar links out to every sub-practice and every legitimately-served city page. Anchor text carries the priority keyword, not “click here.” The linking is programmatic, not manual, because at fifty pages you can’t do it by hand and stay consistent.

Most legal SEO agencies stop at “link to your practice area pages” because the next step needs a developer. Our internal linking strategy work is where a lot of the ranking movement actually comes from, and it’s the step generalist marketing shops usually skip.

Content depth means answering the case type, not hitting a word count

The field’s word count debate (1,200 versus 2,000 versus 2,500) is the wrong argument. The real test is whether the page answers the case type well enough that a prospective client reading it recognizes their situation in the first paragraph.

A good divorce page names the state statute, walks the timeline from filing to decree, explains what happens at the first hearing, and gives an honest fee range. A good custody modification page explains the “material change in circumstances” standard, cites the relevant case law, and describes what evidence courts typically want to see. A good DUI page walks the arraignment, the plea options, the license suspension timeline, and the cost of a diversion program versus trial. Word count follows from the case type. Some questions take 1,400 words to answer honestly. Some take 2,800.

How the content actually gets produced at firm scale, without either bespoke attorney drafting at $2,000 per page or spun garbage: First, a data-driven outline pulled from Google Search Console impressions, SEMrush competitor structure, and the actual questions ranking pages answer. Then AI-assisted first draft against that outline. Then attorney review and rewrite for accuracy, statutory precision, and firm voice. Finally, a QA pass with a tool like Grammarly for clarity and consistency.

We would never auto-publish legal content, and we would never pay $2,000 a page for a paralegal to write a divorce explainer either. The volume gets produced from the outline; the accuracy comes from the attorney review pass. That’s the middle path the SERP refuses to name.

Multi-office firms are a multi-geography indexation problem, not a GBP problem

Claim the Google Business Profile for each staffed office. Keep NAP identical across Avvo, Justia, FindLaw, and Martindale. Respond to reviews. That’s the floor, and it’s the section every ranking piece spends two pages on. Now the actual problem.

Five offices means five GBPs, five city pages, and five canonical decisions, and it’s usually where the indexation falls apart. What goes wrong: two of the city pages canonicalize to each other because the templates were copy-pasted. The Atlanta office GBP points to the Charlotte city page because the URL was pasted wrong on setup. Sub-practice pages exist for the flagship office and nowhere else, so the other four markets are ranking a thin pillar page against firms that have the depth.

The mechanic is the same indexation-and-localization problem we solved for an enterprise client running twenty to thirty top-level domains and country codes: hreflang tags where needed, one canonical target per document, and a clean map of which URL is the source of truth for which geography. The scale is smaller for a five-office firm, but the logic is identical. One GBP per legitimately-staffed office, one canonical city page per office, and sub-practice pages routed to the city page that actually serves that jurisdiction.

Most multi-office legal audit reports we’ve read spend two pages on directory completeness and skip the canonical conflict that’s costing the rankings. The directory work matters. It isn’t the load-bearing work.

Reviews under state bar rules: what we won’t advise on, and where to go instead

Superlatives (“best,” “top-rated,” “highest success rate”) are banned or restricted in many state bar advertising rules. Testimonial disclaimers vary by jurisdiction. Some states require specific language on paid endorsements or prohibit them outright. This is bar advertising law, not SEO, and the firm’s state bar rules are the source of truth. Any SEO agency that promises review request copy without asking which state you practice in is either uninformed or hoping you are. We handle review acquisition cadence, response protocols, and schema markup for reviews where the state permits it. We don’t write review request copy that touches restricted language, and neither should anyone else you hire.

Family law, PI, and criminal defense: intent handling for high-emotion queries

A parent searching “emergency custody modification Atlanta” at 11 p.m. is not the same audience as a B2B buyer researching CRM software. The searcher is scared, tired, and skimming.

Content built for that reader has to acknowledge the situation in the first sentence, put the direct answer to the case-type question in the first 80 words, and defer the firm’s sell to below the fold. Family law, personal injury, criminal defense, and immigration all carry versions of the same emotional load, and the intent-handling moves are the same across all four.

What doesn’t work: keyword-stuffed intros (“Looking for a family law attorney in Atlanta? Our experienced family law attorneys…”), aggressive CTAs above the fold, and phrases like “aggressive representation” or “we’ll fight for you.” The field’s family law templates read like divorce is a product category. The reader can tell in the first sentence, and they close the tab.

What works: a plain first sentence that names what the reader is dealing with, a direct answer to the case-type question (“In Georgia, emergency custody modifications require immediate danger to the child; here’s what that means and what a judge will want to see”), then the depth. The firm’s phone number is at the top of the page. It doesn’t need to be in the copy. Our family law SEO work starts here, because the intent handling has to be right before the architecture matters.

E-E-A-T for legal: named authorship, bar credentials, and the schema that carries the citation

Google’s quality guidelines lean heavily on Experience, Expertise, Authoritativeness, and Trustworthiness for YMYL content, and legal sits squarely inside that classification. In practice, E-E-A-T ships as two things: named authorship on every substantive page, and schema markup that resolves the author and the firm as real entities.

Attorney bios belong in the page, not just in the footer

Every practice area page and every substantive informational page carries a named author or reviewer, in-body, with bar admission jurisdictions, years practicing, and case types handled. Not just a byline. A short block that says “Reviewed by [Name], admitted to practice in Georgia and Florida, twelve years handling custody modifications.” The bio page itself should be linked from the byline and should carry the same credential detail plus a link to the state bar profile. This is what Google’s raters are checking for, and it’s what earns the trust of a reader who is about to hand a firm several thousand dollars.

Person, Attorney, and LegalService schema: the fields that matter

Most legal sites have Attorney schema. Most have it wrong. The sameAs links to bar association profiles, LinkedIn, and Justia are the ones doing the entity resolution work, and they’re the ones almost always missing. Person schema should carry worksFor, alumniOf, knowsAbout, and sameAs. Attorney schema on bio pages. LegalService schema on practice area pages naming the jurisdictions actually served. Article schema on informational content with the author field populated with the Person, not the site name. The leaked Google documentation suggests entity and author signals feed how content is evaluated for both traditional ranking and AI Overview citation, so this is the layer that connects “we wrote a good page” to “Google knows who wrote it.”

What earns AI Overview citations for legal queries

The state of play, from a well-known citation autopsy of legal AI Overview results: Google pulled from state bar association pages, legal marketing agency articles, and YouTube videos. It largely didn’t pull from firm sites. Firms are getting citation-locked out of AI Overviews on their own practice area queries.

Why: firm practice area pages tend to open with a sales pitch instead of a direct-answer capsule. AI Overviews want a paragraph they can extract cleanly. A page that opens with “Our experienced attorneys have decades of combined experience” doesn’t give the model anything to lift. A page that opens with “In Georgia, a divorce becomes final 31 days after the settlement agreement is signed, unless contested; here’s what happens in each of those days” does.

The mechanics that earn citations, in practice: a 40- to 60-word answer capsule under each H2 that could stand alone if pulled; a dated “last updated” line at the top of the page; cited statutes with jurisdiction named; named authorship with schema; and a canonical URL that matches what gets cited. Structured data (LegalService, Article, FAQPage) that clarifies what the page is.

Google’s AI Overview appears in roughly 30% of results and 74% of problem-solving queries, and legal queries skew toward the problem-solving end. The fix for most firm pages isn’t a rewrite. It’s a paragraph, placed in the right spot, that gives the model something to cite.

From ranking to signed case: the attribution stack that matters

Rankings don’t pay retainers. Signed cases do, and the gap between the two is where most legal SEO reporting quietly fails.

The stack, in order: GA4 page view fires on the landing page. A dynamic call tracking number, unique per landing page or per channel, captures the phone call and passes source data. A hidden form field on the intake form captures source_url and landing_page at submission. That data writes to a locked source field in the CRM (Clio, MyCase, whatever the firm uses) that the intake team cannot overwrite. When the client signs, the CRM record carries case_value. Fire generate_lead on form submit and case_value on the signed retainer, not on the phone call.

What most firms have instead: one call tracking number for the whole site, no source field on the intake form, and an intake team that types “referral” into the source field on every case because it’s the fastest option. Neither of those is an SEO problem, but both make the SEO impossible to prove. When we were working with a large financial services client on their debt consolidation and credit card query rankings, the on-page and internal linking work moved the numbers; what made the ROI defensible was the conversion tracking discipline underneath it. Same idea, different vertical.

Fix the intake source field before the third month of the engagement. Otherwise the retainer conversation at month twelve is going to be about clicks, and clicks don’t pay retainers.

What legal SEO costs, and how signed-case ROI actually maths out

The competitor pricing floor, roughly consistent across the SERP: solo firms $1,500 to $3,000 per month, mid-size firms $3,000 to $6,000, multi-office firms $6,000 to $10,000 or more. The billable-rate question the searcher keeps asking (“is $900 an hour a lot”) is a legal-services question, not an SEO one, and I do not have an opinion on it.

Here is where I sit in that range. I start most firms with a one time SEO Diagnostic at $1,500, credited toward your first month if you go on to a monthly program. My monthly programs begin at $2,500 per month for a single location in a category that is not saturated. A law firm competing in a major metro starts at my Growth tier, $4,000 per month, which includes everything in the $2,500 tier, technical upkeep and local search, plus content strategy and production, internal linking architecture, structured data and schema, and AI search visibility, implemented on the site rather than handed over as recommendations. Multi-office or multi-practice firms sit in my Competitive tier at $5,000 to $6,000 per month. CRO, UX and development can be added to any tier for $1,000 per month, and the full list is on my SEO pricing page.

The ROI math, though, is where the field misleads. Monthly retainer against monthly clicks is the wrong denominator. The right denominator is annual retainer against the average signed-case value in the practice areas being targeted. One personal injury case at $40,000 in fees justifies more than a year of a mid-size retainer. One family law retainer at $8,000 covers two months. Estate planning at $2,000 per case needs referral overflow and volume to justify heavy organic investment, and for a lot of solo estate planners in saturated markets, SEO is the wrong channel entirely.

The B2B SaaS clients we’ve worked with tend to see signed-outcome ROI in the multiples of the retainer once the compound effect kicks in, and legal follows the same pattern for the same reason: content and architecture keep earning after the retainer month closes. The number that matters is case value in the practice area being targeted. Anything else is agency math dressed up as strategy.

The first 90 days versus the first year: what actually gets fixed when

Phase 1 (0 to 90 days): blockers only. Indexation audit against the sitemap and Google Search Console coverage report. Canonical conflicts on practice area and city pages resolved. GBP completeness confirmed for every legitimately-staffed office. Core Web Vitals brought to threshold (LCP under 2.5 seconds is the one that usually breaks first on WordPress firm sites). Thin practice area pages either differentiated with real content or noindexed until they can be. Nothing pretty. Nothing that requires a design refresh. Just the blockers.

Phase 2 (90 to 270 days): architecture and depth. Sub-practice pages built out for the practice areas the 80/20 spreadsheet identified. Dynamic internal linking implemented (this is where the developer earns their invoice). Named authorship and Person, Attorney, and LegalService schema shipped across practice area and bio pages. City pages differentiated with real jurisdictional content: courthouses, judges, filing quirks, verdict data. This is where the rankings start moving, and it’s also where the impatience usually sets in. Hold the plan.

Phase 3 (270 to 365 days): attribution stack live and reporting against signed cases, not clicks. AI Overview readiness across the top-priority pages (answer capsules, dated updates, schema audit). Freshness cadence set: personal injury and criminal defense every six months, immigration quarterly, family and estate annually. The site is now doing the work it should have been doing at day one.

Most legal audit reports try to ship all three phases in month one and end up shipping nothing. The 90-day rule for us is: touch nothing that isn’t a blocker.

Questions law firms ask before hiring an SEO

Does SEO work for law firms?

Yes, with conditions. It depends on market saturation, the retainer floor the firm can sustain for eighteen months, timeline expectations that respect the 60- to 120-day floor to movement, and whether the target practice area has enough search demand to matter. The failure modes named earlier in the piece are the honest ones.

What is the 80/20 rule for lawyers?

In SEO terms, two or three practice-area/market pairs typically drive the majority of signed cases from organic, and roughly half the site’s URLs either shouldn’t exist or should support the money pages instead of competing with them. The map is built from Google Search Console impressions, SEMrush competitor gaps, and the firm’s own case-value data.

Is $900 an hour a lot for a lawyer?

That’s a legal-services billing question and outside our lane. What the number tells us about SEO ROI: a firm charging that rate is likely targeting cases where average matter value runs in the tens of thousands, which means one signed case a quarter from organic typically clears a full year of even a top-end retainer. The math favors organic in high-value practice areas.

What are the top 5 SEO strategies for a law firm?

Distilled honestly: first, practice area architecture with the pillar, sub-practice, and city tiers as separate documents. Second, dynamic internal linking wired to route authority into the money pages. Third, local signals (GBP, canonical city pages, differentiated jurisdictional content) for the markets the firm actually serves. Fourth, named attorney authorship with Person, Attorney, and LegalService schema populated correctly. Fifth, answer-ready content structured for AI Overview citation. That’s the list. Everything else is downstream of those five.

If your practice area pages aren’t ranking, start with the architecture

Do the architecture review before the content rewrite. In our engagement with Davis Adams, the firm saw meaningful growth across the target practice areas during the engagement period; we’re not claiming causation for every uptick, but the technical and architectural work is where we spent the time, and the pages that got the work are the ones that moved.

We won’t promise position one, we won’t promise a 3.4x lift in consultations, and we won’t tell you the review is free. What we will tell you is which two or three things in your architecture are the reason your practice area pages aren’t ranking. Request a firm visibility review if that’s the conversation you want to have.