Vertical Playbooks · established evidence
One Verifiable Identity: Why Your Bar Profile Is the Record an AI Engine Actually Trusts
A generative engine deciding which firm to name in an answer is doing something narrower than ranking a page, it is trying to resolve a real-world entity it can trust. That resolution depends on the firm's core facts, name, address, phone, and for a licensed profession, credential data, reading the same way everywhere the engine can check them: the firm's own site, its Google Business Profile, the legal directories, and the state bar record. When those facts disagree, the engine has less confidence in the entity, and less confidence usually means the firm is left out rather than guessed at. The state bar directory carries particular weight here, because it is the one record that actually proves licensure, and current 2026 schema guidance has also moved firms off an older markup type, adding a second, more technical version of the same consistency problem.
Why entity resolution matters more than keyword matching now
Classic search ranking is, at its core, a matching problem: does this page contain the words and signals relevant to this query. A generative answer engine is doing something closer to a verification problem: having found several candidate businesses, which one can it state with confidence is the real, current, correctly described entity behind the answer it is about to give. That verification leans on entity resolution, matching scattered mentions of a name, address, and phone across the web back to one confirmed real-world business.
For a law firm specifically, this intersects directly with licensure. An engine naming a firm in answer to "best injury lawyer near me" is implicitly vouching for a licensed professional, which raises the bar for how confidently it needs to resolve the entity before it is willing to name it at all.
Where the inconsistency actually lives
In practice, a firm's core facts frequently drift across the places an engine can check them: the firm's own site might list a suite number a listing dropped, a name might appear with or without "LLC" depending on the source, a phone number might be a legacy line still indexed on an old directory, and attorney bar-admission details are often simply absent from most listings even where the firm name is correct. Each individual discrepancy looks minor. Together, they are exactly the pattern that makes an engine less certain it has correctly identified the entity behind any one of them.
The legal directories carry unusual weight in this specific resolution problem compared to a typical local business. Justia, Avvo, Martindale-Hubbell, and FindLaw function as a de facto secondary verification layer for the profession, and the state bar directory sits above all of them as the primary licensure-verification source, the one record that actually confirms a named attorney is admitted and in good standing. An engine that can cross-reference a firm's claimed identity against a matching, current bar record has a materially stronger basis for trusting the rest of what it finds.
The technical layer: schema has moved, and a lot of firm sites have not
On top of the directory-consistency problem sits a more technical one. Schema.org, the shared vocabulary sites use to mark up structured data for search engines, has deprecated the dedicated Attorney markup type that many legal sites were built around. Current 2026 guidance directs firms toward LegalService to describe the firm itself and Person, carrying bar-admission detail, for each individual attorney. This guidance is sourced from industry practitioner guides rather than independently re-verified against Schema.org's own changelog in this research, so it is tiered emerging, but the practical implication holds either way: markup built to the old standard is both technically stale and a comparatively weaker signal for how a modern engine reads the page, on top of whatever off-page inconsistency already exists.
Neither problem, the directory drift or the stale schema, is exotic to fix. Both are matters of establishing one canonical version of the firm's facts and correcting every place that disagrees with it, on-page and off. What makes it worth doing deliberately, rather than assuming it is fine because "the phone number is basically right everywhere," is that the cost of an unresolved entity is not a ranking penalty a firm can see, it is simply being left out of the answer with no notice given.
The evidence
Key findings, with their sources
-
Google Business Profile signals account for roughly 32% of local-pack ranking weight, with a distinct, citation and entity-dominated AI Search Visibility ranking category new for 2026.
established Whitespark, Local Search Ranking Factors, 2026 edition.
-
Adding cited statistics, quotations, and authoritative sources measurably raised a source's visibility inside generated answers in tested engines, direction applicable to entity and content signals broadly.
established Aggarwal et al., GEO: Generative Engine Optimization, KDD 2024, arXiv:2311.09735 (peer-reviewed).
-
Schema.org has deprecated the dedicated Attorney markup type; current 2026 guidance directs firms to LegalService (the entity) and Person (each attorney, carrying bar admissions).
emerging Growlaw / PexNet, 2026 legal schema guides.
Reference
Glossary
- Entity resolution
- The process by which a search or AI system matches multiple, scattered mentions of a name, address, and other identifying facts back to one confirmed real-world business or person.
- NAP consistency
- Shorthand for name, address, and phone consistency, the practice of ensuring a business's core facts read identically everywhere they appear online.
- LegalService / Person schema
- The current Schema.org structured-data types recommended for law firms: LegalService describes the firm as an entity, and Person, applied to each attorney, can carry bar-admission and credential detail.
Straight answers
Frequently asked questions
What is the single highest-impact fix for a firm with scattered listings?
Establishing one canonical, written record of the firm's core facts, legal name, address exactly as it should read, primary phone, and attorney bar numbers, and then correcting every directory, profile, and page that disagrees with it against that one source, rather than fixing listings one at a time from memory.
Does the state bar directory actually affect search or AI visibility, or is it just a compliance requirement?
Both. It is required for licensure regardless of marketing, but it also functions as one of the strongest available third-party verification sources for a firm's identity, which is precisely the kind of corroborating signal an engine deciding whether to trust and cite a firm is looking for.
How urgent is fixing outdated Attorney schema specifically?
It is a real but secondary fix behind directory and profile consistency, worth correcting as part of the same pass rather than as an emergency on its own. The guidance behind the deprecation is tiered emerging in this piece because it comes from industry practitioner sources, not a direct Schema.org changelog review, so it is fair to prioritize the off-page identity work first and treat schema correction as part of the same technical pass.
Can fixing directory listings alone get a firm cited in AI answers?
It removes one specific barrier, not the whole gap. A resolved, trustworthy entity is closer to being citable, but an engine still needs relevant, specific content to cite for a given question, which is a separate, content-side problem from identity consistency.
Provenance
Sources
- Whitespark, Local Search Ranking Factors, 2026 edition (established)
- Aggarwal et al., GEO: Generative Engine Optimization, KDD 2024, arXiv:2311.09735 (peer-reviewed, established)arxiv.org
- Growlaw / PexNet, 2026 legal schema guides (emerging)
- Clio, Best Lawyer Directories, and LawSEO, Importance of Legal Directories for Law Firms, 2026 (industry, directional)
Every figure above is attributed to a real, dated source and tagged with its evidence tier. Where a claim could not be verified to a primary source, it is not stated as fact.