Your License Number Is Public Record — AI Search Still Can't Find It On Your Own Site
Ask ChatGPT or Perplexity "licensed electrician in Torrance" and watch what happens. You get a short list, usually pulled from Angi, Yelp, or a local directory aggregator — rarely the electrician's own domain, even when that electrician has held a C-10 license for fifteen years, carries a $1 million liability policy, and has a truck sitting in a driveway two blocks from the person asking. The business is real, licensed, insured, and completely invisible to the machine doing the recommending.
This isn't a ranking problem in the traditional sense. It's a verification problem. And it's showing up across plumbing, electrical, HVAC, and general contracting sites throughout the South Bay right now, quietly redirecting leads to third parties who never touched the job.
Your "Licensed & Insured" Badge Is a Picture, Not a Fact
Walk through twenty trade websites in Torrance, Redondo Beach, or Gardena and you'll find the same pattern. Somewhere in the footer sits a small badge — a shield icon, sometimes a state seal graphic — reading "Licensed, Bonded & Insured." It looks credible to a human visitor. It is functionally invisible to everything else.
Search engines and AI models don't parse images for meaning. They parse text and structured data. A badge that says "Licensed" without a license number, issuing body, and classification next to it in actual HTML text is a decorative claim, not a verifiable one. It's the equivalent of a resume that says "Trustworthy" in the header with nothing to back it up.
Compare that to what shows up on Angi, Yelp, or the CSLB's own contractor lookup tool: a specific license number, a classification code (C-36 for plumbing, C-10 for electrical, C-27 for landscaping, B for general building), a status ("Active"), and an issue date. That's a fact. It's specific, checkable, and formatted the same way every time. When a large language model is deciding which source to cite for "is this contractor legitimate," it isn't choosing between two equally credible sources — it's choosing between one source that hands it a verifiable data point and another that hands it a stock photo of a badge.
Why AI Search Rewards the Aggregator Over the Business Itself
Here's the mechanism, and it's worth understanding because it's not going away — it's getting stronger as more search volume shifts from ten blue links to a single synthesized answer.
Answer engines build what's effectively a confidence score for every claim they might repeat to a user. A claim backed by structured, matchable data — a license number that appears identically on your site, your Google Business Profile, and the state licensing board's own database — scores high. A claim that only exists as unstructured marketing copy ("we're fully licensed and insured!") scores low, because there's nothing to cross-reference it against.
Directories win this game almost by accident. Angi and Yelp normalize license data into structured fields because their business model depends on it — it's how they get paid for lead generation and how they defend against fraud claims. Your own website, built off a WordPress template three years ago by whoever was cheapest at the time, has no such field. The license number lives on a truck decal and a permit application, not in a crawlable, schema-tagged block of text on your own domain.
The result: the AI model has more machine-readable trust signal about your business coming from a $40/month Yelp listing than from the website you actually own and pay to maintain. It cites what it can verify. Right now, that's not you.
What a Verifiable Contractor Site Actually Looks Like
Fixing this isn't a redesign — it's an entity-data problem, and it has a specific shape.
License information belongs in text, not just imagery. Your About page or a dedicated Trust/Credentials page should state, in plain sentences, your license number, classification, issuing date, and status — exactly as it appears on the CSLB record. "License #987654, California C-36 Plumbing Contractor, active since 2011" is a sentence a language model can lift and repeat with confidence. A badge graphic is not.
Structured data needs to carry the same claim in a machine-readable format. Schema.org supports hasCredential, pointing to an EducationalOccupationalCredential object with credentialCategory set to "license" and an identifier matching your state license number. This is the layer most contractor sites skip entirely — not because it's hard, but because whoever built the site never heard of it. It's a handful of lines of JSON-LD, and it's the difference between a claim and a verified claim in the eyes of the system deciding who to cite.
Every instance of the number has to match exactly. Your website, your Google Business Profile (which has a dedicated license field most owners never fill in), your footer, and the CSLB database all need to show the identical number, format, and business name. Inconsistency here doesn't just look sloppy — it breaks the cross-reference that gives the claim its credibility in the first place.
Link out to the primary source, not just around it. Counterintuitively, linking directly to your CSLB verification page from your own site strengthens your position. You're not hiding the check — you're inviting it, on your own turf, which is a stronger trust signal than hoping nobody looks.
Insurance and bonding get the same treatment. A general liability policy number, a bond number, workers' comp status — these are the same category of fact as the license itself, and they should live in the same structured, text-based, cross-referenced format. A general contractor pulling multiple permits a month has more of these credentials than a solo electrician, and each one is a separate opportunity to be the verified answer instead of the unverified one.
None of this replaces reviews, response speed, or the quality of the actual work. It sits underneath all of that — it's the layer that determines whether an AI system trusts your site enough to use it as a source at all.
The Businesses Winning This Right Now Aren't Bigger — They're More Specific
The South Bay contractors starting to show up in AI Overviews and ChatGPT recommendations aren't the largest outfits. They're the ones who treated their license number as a data asset instead of a compliance checkbox. It costs nothing to publish. It takes an afternoon to structure correctly. And almost nobody in the trades has done it, which means the ones who do get an advantage that has nothing to do with ad spend or review count.
The CSLB database isn't going anywhere, and neither is the shift toward AI-mediated search. The question is whether your own domain becomes a source the machines trust, or stays a page they skip on the way to citing Angi instead. If you want a specific read on how your site currently verifies — or fails to — talk to us about what serving contractors across the South Bay has taught us about closing this gap before your competitor does.