The American sector furthest behind on AI search visibility, and the cheapest way into it
Why professional and local services are the gap, what the citation evidence says about who the engines trust instead, and the three assets that close it fastest.
Published by AI Knows Us · Updated 22 October 2026
The American sectors furthest behind are professional services and local services, and the evidence for that is not a survey. It is a citation list. In 64 blind readings of ChatGPT and Gemini on American buying questions on 22 October 2026, the engines named software products 48 times and named not one American services firm of any kind. Software vendors publish specifications, prices, documentation and changelogs, all dated and checkable. Professional and local services firms publish claims. The engines quoted the first and ignored the second, which makes the gap a publishing gap rather than a budget gap, and that is good news because publishing gaps are cheap to close.
Why services sectors lose this queue
| What the engines quoted | What software publishes | What most services firms publish |
|---|---|---|
| A price | A number, on a page, with tiers | Contact us for a quote |
| A specification | What is included, counted | A description of an approach |
| A date | Changelogs and updated notices | Undated pages |
| A denominator | Usage numbers with a base | Percentages without one |
| Documentation | Reference material a third party can check | Case studies written by the seller |
Read down the right hand column and the reason for 53 namings going to software becomes mechanical rather than mysterious. An engine composing an answer reaches for the thing it can quote with attribution.
The three assets that close the gap fastest
- A price position, published and dated. Not a full rate card if that is commercially impossible, but a starting figure, a typical engagement range, or a clear statement of what is quote only and why. A page that says our smallest engagement is X dollars and here is what it includes, read on this date, is quotable. Contact us is not.
- One original dated figure about your own market. Something you counted, with a denominator and a method. Services firms sit on this data and almost never publish it. It is the single most quotable asset available to a professional practice and it costs internal time rather than money.
- A specification of the service itself. What arrives, how many of it, in what format, by when, and what is excluded. Written as a specification rather than as a promise. This is the document procurement wants and the document an engine can cite.
The constraint services firms actually face, and how to work inside it
Professional practices in the United States operate under conduct rules that govern advertising, and those rules are real. State bar advertising rules restrict comparative and outcome claims for law firms. Financial advisers are subject to the SEC marketing rule and to FINRA requirements on communications. Healthcare marketing touches patient privacy obligations. Over all of it sits Section 5 of the FTC Act and the FTC Endorsement Guides, with the Commission's 2024 rule on consumer reviews making undisclosed insider and incentivised reviews directly actionable.
None of those rules prohibits publishing a price, a specification or a counted figure about your own market. They restrict outcome promises, testimonials without disclosure, and comparative superiority claims. So the three assets above are available inside the constraint, and a practice that thinks it cannot publish anything has usually confused a rule about claims with a rule about facts. The reviewer who clears each asset is the same reviewer the firm already needs.
What the readings showed in detail
On 22 October 2026 we put a 198 question American buyer bank in 33 batches to the ChatGPT and Gemini product surfaces through a data provider, with the United States named inside the text of every question, because the provider accepts gl, hl and location parameters and ignores all three. Batch 33 names us in the question, so it measures a ceiling and is excluded from every figure.
- 64 blind readings, 32 from ChatGPT and 32 from Gemini.
- 2,866 citation URLs across 1,143 distinct hosts.
- Named us 0 times, cited our domain 0 times.
- Named a rival in 24 readings, named nobody at all in 38.
- 53 namings, every one of them software. Semrush 14, HubSpot 12, Ahrefs 9, BrightLocal 5, Peec 3, Profound 2, SE Ranking 2, Otterly 2, Scrunch 1, Surfer 2, Conductor 1.
- Top hosts, developers.google.com 17, support.google.com 15, searchengineland.com 13, blog.google 11, clutch.co 11, semrush.com 10, ftc.gov 9, blog.hubspot.com 9. The entry ccn.web.tr at 8 is the provider's own fetch infrastructure rather than a publisher and is excluded.
ChatGPT ran a live search on all 32 of its American readings and Gemini on 16 of its 32, so the two surfaces are reported separately throughout and never averaged.
The first ninety days for a services firm
- Days 1 to 10. Take a baseline. Twelve buying questions in your category, both engines, United States named in the question, answers and sources saved and dated.
- Days 11 to 30. Publish the price position, cleared by whoever clears your claims. Date it.
- Days 31 to 60. Publish the service specification. What arrives, how much, by when, what is excluded.
- Days 61 to 80. Count one thing about your own market and publish it with its denominator, its method and its date.
- Days 81 to 90. Re-read the same twelve questions, same engines, same denominator, and report the two engines separately.
If your practice is one where the conduct rules shape every asset, the sequencing for that is worked through in why the word professional changes the brief. If you want to see what a specification shaped page looks like when a vendor publishes one, the audit of twelve pricing documents in twelve pricing documents audited is the pattern to copy.
Questions about which American sectors are behind
Why are software vendors ahead?
Because the things they already publish to sell software, prices, specifications, documentation and changelogs, are exactly the things an engine can quote with attribution. It is a by product of how they sell rather than a visibility strategy.
Can a law firm publish a price?
Fee information is widely published by American firms in some practice areas and the conduct rules restrict outcome claims rather than fee disclosure. Clear it with whoever clears your advertising and publish what you can, dated.
Is this just content marketing?
The assets are different. A price position, a specification and a counted figure are reference documents rather than articles, and on our reading reference documents are what the engines quoted.
Which single asset should come first?
The price position, because it is the shortest to write and the one most often missing. The counted figure has the highest ceiling but takes the longest.
How long before the reading changes?
Treat ninety days as the first fair re-read, and remember that on our evidence one engine read the live web on every American reading and the other on half of them, so the two will reflect a change at different speeds.
Measuring a gap before you spend on closing it
The claim that a sector is behind is only useful to a firm that knows where it personally stands, and on 22 October 2026 our American readings named no services firm at all in 64 readings. AI Knows Us reads ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews for the buying questions in a category, reports each engine separately with its denominator, and keeps the raw answer text so the before and after are comparable rather than asserted. In the United States that is the Growth plan at 299 dollars a month with 100 buyer questions watched and 1,600 credits, read on aiknowsus.com/pricing and aiknowsus.com/llms.txt on 22 October 2026. The 149 dollar SEO Start tier for the United States and the United Kingdom reads zero engines and would not produce this reading.