Ask ChatGPT for the best HVAC company in Kansas City. Then ask Claude. Then ask Perplexity.

Watch what happens. Sometimes you get a great local KC shop. Sometimes you get a national name like ServiceMaster or one of the big franchises. Sometimes the three platforms give you three different answers.

Run the same test for plumbing. For pest control. For restoration. For roofing.

Here is the pattern that holds: the businesses AI cites consistently and describes accurately are the ones with strong signals. The local shop that AI sometimes recommends and sometimes forgets is the one with weak signals. The national names that show up everywhere are the ones with twenty-five years of compounding signal in the systems AI was built on and the live web AI pulls from.

This is not because anyone bought their way in. AI does not have ad slots. ChatGPT does not sell sponsorships. There is no “promoted business” tag inside Perplexity.

So what is actually happening?

Signals.

ServiceMaster has spent 25 years putting signals into the world that AI was eventually trained on. Most local businesses have never built those signals because nobody told them they mattered, and back then they didn’t.

Here is what ServiceMaster actually has that small local HVAC businesses do not:

A Wikipedia entry. Wikipedia is one of the most heavily weighted training sources for every major LLM. Your local HVAC shop does not have a Wikipedia entry. ServiceMaster does. AI was trained on Wikipedia.

AI now knows ServiceMaster exists.

A Wikidata entity. Wikidata is the structured-data backbone of the open web. ServiceMaster has a Wikidata entry with their address, founding date, locations, industry codes, parent company, and a list of services. The crawler that fed AI knew exactly what ServiceMaster was. Your local shop’s Wikidata entry, if it exists at all, has a name and a website. That is the entire signal.

Schema markup on every page. Their site has LocalBusiness schema, Service schema, FAQPage schema, and Organization schema. Every page tells AI what the page is about in machine-readable JSON. Your local shop’s site probably has zero schema. AI has no idea what’s on your pages.

Consistent name, address, and phone across hundreds of third-party sites. ServiceMaster appears identically on Yelp, Angi, HomeAdvisor, the BBB, every chamber of commerce in every metro they operate in, and a hundred industry directories. AI saw the same business name and address 800 times. It learned the entity.

A decade of press mentions. Local papers, trade publications, industry awards. Every mention added to the citation graph AI was trained on. Your local HVAC has maybe one press mention from when they opened.

Structured FAQ content on their site. Every page has questions and answers in a format AI can pull and quote directly. The chunks of text most likely to be cited verbatim by ChatGPT are FAQ answers, because they read like AI responses already.

Twenty-five years of compounding signal. ServiceMaster has been generating press mentions, directory listings, and Wikipedia-adjacent content since the late 90s. Most of those signals ended up in the training data AI was originally built on, AND they show up across the live web that today’s AI platforms search in real time. AI sees their name everywhere, in every input it touches. Your local HVAC has been doing the right work for their customers for years. They just haven’t been doing the work that gets them into AI’s mouth across both training data and live search.

None of those signals require a $30,000 a month ad budget. They require structure, methodology, and consistent execution.

Here is the punch line: every signal ServiceMaster has, a local business can have. The signals are not impossible to build. They are not even particularly expensive to build. What they require is methodology, specialized technical work, and constant maintenance because what AI rewards changes month to month. This is not a weekend project. This is not a one-time install. This is an ongoing service motion that nobody in the small business marketing world has been running because nobody was selling it. Until now.

You don’t need ServiceMaster’s budget. You need their signals.

The reason small businesses don’t have these signals is not money. It’s that nobody told them the signals existed. SEO agencies sold them on backlinks and keyword density. Those signals do not move AI citation rates the way agencies promised. AI synthesizes from a different set of inputs than Google’s blue-link algorithm does. It cares about Wikipedia, Wikidata, schema, structured content, third-party authority sites it already recognizes, and the live web data it pulls in real time. Backlink graphs and keyword density are a smaller piece of that picture than the agencies who sold them led businesses to believe.

That is the gap. That is what Discover Me deploys.

If you want to see where you stand on those signals specifically, my free scan at https://scan.leftymediaco.com runs a snapshot. Takes 30 seconds.