Entity Consistency for Law Firms: The One Fix Most Firms Skip

Short answer: An entity is the thing a search system believes exists: a firm with a name, an address, a phone, a set of lawyers and a set of practice areas. The system builds that belief from every place the firm is mentioned and trusts it in proportion to how well the mentions agree. Inconsistent mentions split the firm into several weak entities and make AI answers hesitate to name it. The fix is unglamorous: one canonical set of facts, applied character for character to the site, the schema, the bar profiles and the directories, then monitored.

What is an entity, in practice?

Google's Knowledge Graph and the models behind AI answers do not index a firm; they index statements about it. "Smith & Lee is a family law firm at 500 N Brand Blvd, Glendale, phone (818) 555-0100, with attorneys Jane Smith and David Lee." When the site, the State Bar, Avvo, Justia, Yelp, the county bar and the schema markup all make that statement the same way, the system holds it with confidence. When they disagree, it holds several partial versions and cites none of them.

This has always mattered for the map pack under the name of "NAP consistency". It matters more now because AI answers name entities, and an answer that is unsure which of three similar-looking firms is real names a competitor instead.

Where do mismatches come from?

From time, mostly. Firms rename when partners join or leave, move offices, change phone systems, and drop practice areas, and each event updates some listings and not others. The usual sources:

How do you fix it?

Day 1: decide the canonical facts. One document. Legal name exactly as on the State Bar record and the Secretary of State filing. One address format. The main phone. The canonical URL. The current attorney list. The current practice list.

Day 2: the site. Footer, contact page, about page, each attorney profile, and the LegalService and Person schema. Every one matches the document. This is also the moment to check that the schema @id values are stable and that sameAs on the organization node points at the firm's profiles, not at the lawyers' personal ones (see Person schema done right).

Day 3: the source of truth. The State Bar profile for each attorney. Address, phone, website. Directories copy from here.

Days 4–5: the twelve listings. Google Business Profile, Avvo, Justia, FindLaw, Yelp, Apple Business Connect, Bing Places, Martindale, Super Lawyers if present, the county bar, BBB, and any practice-specific directory (AILA, NAELA). Claim, correct, remove duplicates. The directory audit walks through each.

Day 6: the rest. Social profiles, the firm's Wikipedia or Wikidata entry if any, press mentions that can be corrected, the email signature block.

Day 7: monitoring. A quarterly re-check, because scrapers reintroduce old data and new directories appear.

How do you know it worked?

Three signals, over two or three months:

• The Google knowledge panel for the firm name (if one appears) shows the correct facts and stops showing a "suggest an edit" oddity like a wrong category.
• Map-pack position for the core query in the office's city improves or stabilizes; this is the most reliable short-term sign.
• AI answers that mention the firm use the canonical name. We check this in the monthly query run.

What does not work?

• Fixing only Google. Google reads the others.
• Different phone numbers per directory for attribution. Attribution at the cost of consistency is a bad trade; track calls on the site instead.
• Leaving unclaimed directory profiles because "we don't use that site". The system uses it.
• Doing it once. It drifts.

Entity consistency is the first week of every engagement in the law firm SEO service, and the free visibility check reports the mismatches it finds across the listings before anyone signs anything. Why it feeds AI citations specifically is covered in the GEO for law firms guide.