ChatGPT Does Name Agencies
When someone types "recommend a home care agency in Jacksonville FL" into ChatGPT, it will sometimes respond with specific agency names. Not always. Not in every market. But it happens — and the agencies it names are not random.
In our testing, ChatGPT named specific agencies in markets where those agencies had strong, consistent signals across the web: high review counts, complete directory listings, mentions in local publications, and structured content on their websites that answered home care questions directly.
In markets where individual agency signals were weak, ChatGPT defaulted to recommending search strategies — "search Google for home care agencies near you," "check Care.com" — rather than naming specific providers.
The threshold for getting named is not low. ChatGPT is drawing on indexed data across the entire web. The agencies it names are the ones with the most total signal weight in their market.
What Builds the Signal ChatGPT Uses
ChatGPT does not call a directory or read Google Maps in real time. It synthesizes information from training data and, in its web-browsing mode, from current search results. The signals that make an agency recommendable are the same ones that make it visible anywhere:
Review volume and recency — Agencies with 50+ Google reviews appear in ChatGPT's local recommendations at a substantially higher rate than agencies with fewer than 20. Volume outweighs rating. A 4.5 average with 80 reviews is more likely to be named than a 5.0 with 12.
Directory presence — Care.com, A Place for Mom, Yelp, Healthgrades. ChatGPT's training data includes these directories' content. An agency that appears prominently in multiple directories has multiple entry points into that training data.
FAQ and resource content — Agencies whose websites directly answer home care questions have content that AI models can reference and attribute. An agency with a resource section that answers "what does companion care include" is more likely to be cited than one with only a services page.
Local media mentions — A local news story, a community award, a mention in a local business journal. These carry outsized weight relative to their volume because they are the kinds of authoritative, local-specific references AI models look for when building a picture of a market's reputable businesses.
The Timeline Problem
Here is the frustrating reality: building the signal ChatGPT uses to name agencies takes 12 to 18 months of consistent execution. Review generation, directory optimization, content creation. None of these happen overnight.
The agencies that will be named by ChatGPT in 24 months are the ones building the foundation now. Agencies that wait until AI recommendation becomes a commonly understood marketing channel will be building 24 months behind.
The floor is the same as it always was: complete GBP, consistent citations, reviews that reflect the care you actually provide, content that answers the questions families are actually asking.
What the Near-Term Play Is
In most markets today, AI recommendation of specific agencies is not yet consistent enough to be a primary strategy. The near-term value is in appearing prominently within the directories that AI does cite consistently — Care.com, A Place for Mom, Caring.com — because those platforms appear in AI answers whether or not individual agencies do.
A complete, well-reviewed Care.com profile gets surface area in ChatGPT answers even when the agency is not named directly. Families who see your agency's name on Care.com in a ChatGPT-cited result and then search for you specifically are real leads.
The medium-term play — building the content and review foundation that makes direct AI recommendation possible — runs in parallel.
Source: HCB AI Search Visibility Study — 54 keywords tested across 4 AI platforms, 2026. For the full picture on AI visibility for home care agencies, see our AI Marketing for Home Care service.