What AI assistants cite for local businesses

Ask an AI assistant to recommend a plumber in your city and you'll get a short, confident list of names. Ask it why those names, and the answer gets more interesting — because modern engines increasingly show their receipts. Perplexity cites sources inline. ChatGPT and Claude browse and retrieve. Under every recommendation is a pile of documents the engine decided to trust.

Frontpagely records those cited sources on every measurement we run — which pages, which domains, engine by engine, question by question. This post is about the pattern in that citation data for local businesses, and what to do about it. Fair warning: the pattern is boring. That's the good news, because boring is actionable.

The receipts are the usual suspects

When the question is local — "best HVAC repair near me", "family dentist in Tampa worth calling" — the sources engines cite cluster into a few familiar families:

Maps and profile platforms. Google Business Profile data, Apple Maps, Bing Places. These are structured, machine-readable, and corroborated by millions of users — exactly the kind of source a cautious engine leans on for "does this business actually exist, where, and is it open."

Review directories. Yelp and its sector-specific cousins (Avvo for lawyers, Healthgrades for doctors, Houzz for contractors — whatever canon your industry has). They rank well, they're crawled constantly, and they package name + location + reputation in a format a model can quote without going out on a limb.

"Best of" roundups and local press. Listicles are the fast food of the web, and engines eat there anyway — a "10 best movers in Austin" page answers the question in literally the shape it was asked.

Your own website — when it's quotable. Engines do cite businesses directly, but noticeably more when the site states things plainly: what you do, where, for whom, at what hours, with structured data confirming it.

Notice what's missing: nothing exotic. No secret directory, no "AI optimization" trick. The engines trust what the web has collectively vetted.

Why directories punch above their weight

A language model recommending a business is making a factual claim it can't verify firsthand. So retrieval-backed engines look for corroboration — the same business, same name, same address, same phone, appearing consistently across sources that don't share an owner.

That's why listing consistency is more than old-school SEO hygiene. Disagreement between your profiles — an old address on one directory, a variant business name on another — reads as uncertainty to a machine deciding whether to put its confidence behind you. Agreement reads as ground truth.

It's also why the four highest-leverage listings for most local businesses are the unglamorous ones: Google Business Profile, Yelp, Apple Maps, and Bing Places. Claimed, accurate, and telling the same story. (All four key off your street address — which is why Frontpagely asks for it, and why the listing checks stay parked until you provide one.)

One honesty note, because it's a place tools love to fudge: when Frontpagely checks your presence on these platforms, unclaimed or unverified listings are reported as exactly that — data ready for you to claim — not as "you're all set." A derived check is not a confirmed listing, and we label the difference.

Making your own site worth citing

The directory layer gets you named. Your own site is what gets you quoted. Three things matter, in rough order:

1. Structured data. LocalBusiness schema (or your sector's specific type) states your name, address, phone, hours, and service area in the format machines parse first. Frontpagely generates this prefilled from your business details — it's copy-paste work, not a project.
2. Pages that answer the actual question. If buyers ask "do you handle emergency calls on weekends," a page that plainly answers it is citable; a slogan is not. Write answers, not adjectives.
3. Basic crawlability. Fast, secure, titled, and not accidentally blocking the AI crawlers. Unfixed technical rot quietly disqualifies you from consideration regardless of reputation.

What we won't tell you

No one — including us — can promise an engine will cite you. Anyone who does is selling control they don't have. The engines re-rank, re-crawl, and update without notice.

What a business can control is the inputs: claimed and consistent listings, structured data, answer-shaped pages. What a measurement tool can do is close the loop — record who the engines name and cite for your questions this week, hand you the gap list, then re-measure after you've done the work so you can read the trend next to the completed fixes. That's the whole honest playbook. It's less exciting than a growth hack and it has the advantage of being real.

The free audit shows you today's answer to both questions — who the engines name for your market, and which sources they pulled from — in about a minute.