Research / Local business AI readiness, October 2026
AI's door is open. Local businesses have nothing to say.
Nearly every local service business lets AI in, and few give it what it needs to recommend them. We checked 1,238 local service business websites in 13 trades and 7 US metros between October 5 and 18, 2026. 98.9% let all five major AI crawlers in through robots.txt, but only 7.4% pass all five of our AI-readiness checks. Fewer than half (46.4%) state the business's name, phone, and address in LocalBusiness structured data, 29.2% publish a price, and 2 sites (0.2%) publish a dataset.
By Ben Hinton, founder of CiteMy.ai
, published
Do AI crawlers get into local business websites?
Yes, by the sites’ own rules. 98.9% of the sites (1,224 of 1,238) let all five AI crawlers we checked into the homepage through robots.txt: OpenAI’s GPTBot and OAI-SearchBot, Anthropic’s ClaudeBot, PerplexityBot, and Google-Extended, the token Google reads for AI use. Only 14 sites (1.1%) block any of them: 11 block ClaudeBot, 10 GPTBot, 9 Google-Extended, 2 PerplexityBot, and none blocks OAI-SearchBot, the crawler OpenAI says puts sites in ChatGPT search.
That is the stated rule, not proof of access. A firewall or CDN can turn a crawler away without a word in robots.txt, and 98 sites in our sample refused even our own declared research crawler (they are not in the 1,238). Local businesses are not shutting AI out. The question is what it finds once it is in.
How many local business websites are ready for AI search?
7.4% (92 of 1,238) pass all five of our AI-readiness checks, and 22.9% (284) pass at most one. Most land in the middle: 311 sites pass two checks, 302 pass three, and 249 pass four.
| Check | All (n=1,238) | Home services (n=816) | Professional (n=422) |
|---|---|---|---|
| 1. AI answer-engine crawlers allowed | 98.9% | 98.9% | 98.8% |
| 2. A valid llms.txt | 32.4% | 30.9% | 35.3% |
| 3. Business details in schema | 46.4% | 45.5% | 48.3% |
| 4. Prices, warranties, or specs published | 41.0% | 49.3% | 25.1% |
| 5. Other citable data | 45.0% | 47.7% | 39.8% |
| Passes all five | 7.4% | 9.8% | 2.8% |
| Passes at most one | 22.9% | 22.4% | 23.9% |
| Publishes a price | 29.2% | 31.4% | 25.1% |
| Has a table of numbers | 10.7% | 12.3% | 7.6% |
| Publishes a dataset | 0.2% | 0.2% | 0.0% |
Shares of the sites judged on all five checks. Home services: roofing, HVAC, plumbing, foundation repair, pool service, solar, remodeling, windows and doors, landscaping. Professional: dentists, med spas, personal-injury law, accountants.
Home-services sites pass all five checks more than three times as often as professional-services sites (9.8% vs 2.8%). Most of the gap is check 4: 49.3% of home-services sites publish a price, warranty, or spec, against 25.1% of professional sites. That check was written with contractors in mind (warranty terms and product specs), so the sector gap partly reflects the test.
What are local business websites missing?
Who the business is, in a form a machine reads. 79.9% of the 1,305 homepages we scanned carry some structured data (JSON-LD), but only 46.4% of sites have a LocalBusiness node with the business’s name, phone, and address. The rest leave an AI engine to piece those together from page text.
Prices. 29.2% publish a price anywhere on the pages we read. People ask AI assistants what a job costs; a site with no number gives the assistant nothing of its own to quote.
Data. 10.7% have a table of numbers, and 2 of 1,238 sites (0.2%) publish a dataset.
llms.txt is not the gap. 32.4% already have a valid llms.txt file. Many of these files are generated by SEO plugins rather than written on purpose: the HTTP Archive’s 2025 Web Almanac traces 39.6% of the llms.txt files it found to one plugin.
Which trades are the most AI-ready?
Roofing leads: 15.0% of roofing sites pass all five checks, followed by HVAC and foundation repair. Accountants trail at 1.0%. Personal-injury law firms are a split case: 51.8% have a valid llms.txt and 65.2% pass the citable-data check, the most of any trade, but only 14.3% publish a price, so few pass all five.
| Trade | Sites (n) | AI crawlers allowed | Business schema | Publishes a price | All five checks |
|---|---|---|---|---|---|
| Roofing | 120 | 99.2% | 46.7% | 25.0% | 15.0% |
| HVAC | 96 | 97.9% | 61.5% | 43.8% | 14.6% |
| Foundation repair | 66 | 98.5% | 47.0% | 34.8% | 12.1% |
| Plumbing | 89 | 98.9% | 60.7% | 25.8% | 11.2% |
| Pool service | 91 | 100.0% | 42.9% | 31.9% | 8.8% |
| Solar | 91 | 98.9% | 30.8% | 31.9% | 7.7% |
| Remodeling | 79 | 98.7% | 48.1% | 46.8% | 6.3% |
| Windows and doors | 97 | 99.0% | 34.0% | 22.7% | 6.2% |
| Dentists | 107 | 96.3% | 60.7% | 26.2% | 5.6% |
| Landscaping | 87 | 98.9% | 37.9% | 24.1% | 4.6% |
| Personal-injury law | 112 | 100.0% | 61.6% | 14.3% | 2.7% |
| Med spas | 101 | 99.0% | 38.6% | 37.6% | 2.0% |
| Accountants | 102 | 100.0% | 30.4% | 23.5% | 1.0% |
Sites judged per trade (n). Business schema: LocalBusiness structured data with name, phone, and address on the homepage.
Does AI readiness vary by city?
Hardly. The all-five rate runs from 6.5% to 8.3% across the six metros with enough sites to report. Where a business is matters less than what its website says.
| Metro | Sites (n) | AI crawlers allowed | Business schema | Publishes a price | All five checks |
|---|---|---|---|---|---|
| Houston | 222 | 99.5% | 50.9% | 32.9% | 7.2% |
| Dallas | 218 | 99.5% | 47.7% | 29.8% | 8.3% |
| New York | 200 | 98.0% | 42.0% | 28.5% | 6.5% |
| Phoenix | 199 | 99.0% | 47.7% | 31.2% | 7.5% |
| Los Angeles | 196 | 99.0% | 45.9% | 27.0% | 7.1% |
| Chicago | 184 | 97.8% | 41.8% | 25.5% | 6.5% |
Sites judged per metro (n). Atlanta (19 sites, roofing only) is counted in the totals but not shown on its own.
Is AI crawling of business websites growing?
Yes. Cloudflare reports AI and search crawler traffic up 18% from May 2024 to May 2025, with OpenAI’s GPTBot requests up 305%. In 2025 AI bots made 4.2% of HTML page requests on its network, and crawling triggered by a person asking a chatbot grew more than 15 times. Business use of AI is uneven: the Census Bureau finds 23.8% of employer businesses used AI in the two weeks to September 20, 2026, from 17.2% in construction to 42.7% in professional services.
| Measure | Figure | Period | Source |
|---|---|---|---|
| Growth in AI and search crawler traffic | +18% | May 2024 to May 2025 | Cloudflare, Jul 1, 2025 |
| Growth in GPTBot (OpenAI) requests | +305% | May 2024 to May 2025 | Cloudflare, Jul 1, 2025 |
| AI bots' share of HTML page requests | 4.2% | 2025 (Jan 1 to Dec 2) | Cloudflare, Dec 15, 2025 |
| Growth in AI "user action" crawling (a bot fetching a page because a person asked a chatbot) | More than 15x | 2025 | Cloudflare, Dec 15, 2025 |
| US employer businesses that used AI in a business function | 23.8% | Two weeks to Sep 20, 2026 | U.S. Census Bureau, Oct 8, 2026 |
| Construction businesses that used AI | 17.2% | Two weeks to Sep 20, 2026 | U.S. Census Bureau, Oct 8, 2026 |
| Professional, scientific, and technical services businesses that used AI | 42.7% | Two weeks to Sep 20, 2026 | U.S. Census Bureau, Oct 8, 2026 |
| Arizona businesses that used AI | 29.5% | Two weeks to Sep 20, 2026 | U.S. Census Bureau, Oct 8, 2026 |
Outside figures measure different things on different populations (a network's traffic, all US employer businesses). They are context, not a comparison with our sample.
Key statistics
- 98.9% (1,224 of 1,238) local service business websites let GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, and Google-Extended in through robots.txt (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- 7.4% (92 of 1,238) pass all five AI-readiness checks: crawlers allowed, a valid llms.txt, LocalBusiness schema with name, phone, and address, a published price, warranty, or spec, and two or more kinds of citable data (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- Home-services sites pass all five checks more often than professional-services sites: 9.8% (80 of 816) vs 2.8% (12 of 422) (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- 46.4% (575 of 1,238) have LocalBusiness structured data with the business's name, phone, and address on the homepage (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- 29.2% (362 of 1,238) publish a price on the pages we read (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- 0.2% (2 of 1,238) publish a dataset (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- 22.9% (284 of 1,238) pass at most one of the five checks (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- 32.4% (401 of 1,238) have a valid llms.txt file, many of them generated by SEO plugins (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- By trade, roofing sites pass all five checks most often, 15.0% (18 of 120); accountants least often, 1.0% (1 of 102) (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- 1.1% (14 of 1,238) block at least one of the five AI crawlers in robots.txt; none blocks OAI-SearchBot (CiteMy.ai, 1,238 local service business websites, crawled Oct 5 to 18, 2026).
- AI and search crawler traffic grew 18% from May 2024 to May 2025, and GPTBot requests rose 305% (Cloudflare, July 1, 2025).
- AI bots made 4.2% of HTML page requests in 2025, and crawling triggered by people asking chatbots grew more than 15x (Cloudflare Radar Year in Review, Dec 15, 2025).
- 23.8% of US employer businesses used AI in a business function in the two weeks to Sept 20, 2026: 17.2% in construction, 42.7% in professional services, 29.5% in Arizona (U.S. Census Bureau BTOS, published Oct 8, 2026).
What should a local business publish for AI?
Start with what the checks found missing.
- The business’s name, phone, and address in LocalBusiness structured data on the homepage, matching the page text. Our free schema checker shows what yours says.
- Prices: ranges, starting-at figures, or typical job costs from the business’s own jobs.
- Facts a crawler can cite: a license number, years in business, service-area cities, warranties with their length, and FAQs answered with real numbers.
- Data: a table of prices or job outcomes, with a CSV download and Dataset structured data.
- A robots.txt that keeps the search crawlers (OAI-SearchBot, PerplexityBot, and the like) allowed, and an llms.txt that points to those pages. Check both with our free llms.txt and crawler checker.
That is the work CiteMy.ai does every month for local service businesses, measured weekly in ChatGPT, Claude, Perplexity, and Gemini. See what a client gets and pricing, or get a free audit of your site against these checks.
How we measured
Sample: 1,467 local service businesses from Google Places text searches (about the top 20 results for each trade in each city) across 13 trades (roofing, HVAC, plumbing, foundation repair, pool service, solar, remodeling, windows and doors, landscaping (home services); dentists, med spas, personal-injury law, and accountants (professional services)) and 7 US metros (Atlanta, GA; Chicago, IL; Dallas, TX; Houston, TX; Los Angeles, CA; New York, NY; Phoenix, AZ), deduplicated by place and domain. 1,417 had a website; 1,238 could be judged on all five checks. Not judged: 98 homepages refused our declared crawler (HTTP 403), 29 robots.txt files did not answer, 8 sites disallow our crawler, 30 were not yet checked, and 14 had other load errors. Sites were read 2026-10-05 to 2026-10-18: robots.txt, /llms.txt, the homepage, and a sample of other pages. Every share is a count over the 1,238 sites judged (or over a group's judged sites); groups under 30 sites are suppressed (n only). Aggregates only: no business or website is named.
The five checks, as we ran them:
- AI answer-engine crawlers allowed. robots.txt does not block GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, or Google-Extended from the homepage (no robots.txt at all passes).
- A valid llms.txt. /llms.txt is a text file whose first line is a "# " title, as the llmstxt.org format asks.
- Business details in schema. The homepage's JSON-LD has a LocalBusiness node (or a subtype, such as RoofingContractor or HVACBusiness) that itself carries the business's name, a phone number, and an address.
- Prices, warranties, or specs published. On any page we read, a price (an Offer price, a priceRange with a number, or a dollar figure stated as a range, per unit, installed, or "starting at"), a warranty with its length (in years, or lifetime), or a product spec with its value (U-factor, SHGC, SEER2, AFUE, impact class, wind rating, and the like), in structured data or in the page text.
- Other citable data. At least two of five kinds, on the pages we read: FAQ content (FAQPage schema, or an FAQ page or section with three or more questions); a published dataset (Dataset schema, a /data/ section, or a CSV, JSON, or spreadsheet download); an HTML table with numbers in it (business hours do not count); service-area cities (city location pages, or a service-area page naming three or more of the metro's cities); and a business fact (a contractor license number, or years in business).
Limitations
- The sample is Google's top results for each trade and city, so these are the most visible businesses, not a random sample of all local businesses. Trades and metros were chosen for our own outreach, not to represent the US. A random sample would likely score lower.
- The crawl ran for two weeks (Oct 5 to 18, 2026); a site changed after its check is not reflected.
- robots.txt is not access. A site can block AI crawlers at its firewall or CDN without saying so in robots.txt, so "crawlers allowed" means robots.txt allows them, not that they get in.
- Prices, warranties, specs, and citable data are found by keyword and schema rules on a sample of each site's pages; content behind JavaScript or in PDFs can be missed.
- The five checks are ours, not a standard any AI engine publishes. Google, for one, says AI Overviews need no special markup.
- Trade and metro rows are about 66 to 222 sites each. The trade-by-metro cross-tab (about 20 sites per cell) is not published, and Atlanta (19 sites, roofing only) counts in the totals but has no row of its own.
Questions about AI readiness
Do AI crawlers crawl small business websites?
Almost all small business websites let them. 98.9% (1,224 of 1,238) of the local service business sites we checked in October 2026 allow GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot, and Google-Extended in robots.txt. robots.txt is only the stated rule: a firewall or CDN can still block a crawler. Cloudflare reports AI bots made 4.2% of HTML page requests on its network in 2025.
How many local businesses are ready for AI search?
7.4% (92 of 1,238) of 1,238 local service business websites passed all five of our AI-readiness checks in October 2026. Home-services sites did better (9.8%) than professional-services sites (2.8%), mostly because few law, dental, med spa, and accounting sites publish prices.
What makes a local business website AI-ready?
Our five checks: robots.txt lets the AI crawlers in; a valid llms.txt; LocalBusiness structured data with the business's name, phone, and address; a published price, warranty, or product spec; and at least two kinds of citable data, such as FAQs, a table of numbers, service-area cities, a license number, or years in business. These are our checks, not a standard any AI engine publishes.
Does my business need an llms.txt file?
It helps, but it is rarely what is missing. 32.4% of the sites we checked already have a valid llms.txt, many generated automatically by SEO plugins. The bigger gaps are business details in structured data (46.4% have them), prices (29.2%), and data a crawler can cite (0.2% publish a dataset).
Should a local business block AI crawlers?
Very few do: 1.1% of the sites we checked block any of the five. The crawlers do different jobs. OpenAI says OAI-SearchBot is what shows sites in ChatGPT search, while GPTBot collects training data and can be blocked separately; Perplexity says PerplexityBot surfaces sites in its search results. Blocking a search crawler keeps the business out of that engine's answers.
Which local businesses are the most AI-ready?
By trade, roofing (15.0%), HVAC (14.6%), and foundation repair (12.1%) sites pass all five checks most often; accountants (1.0%), med spas (2.0%), and personal-injury law firms (2.7%) least often. By metro the rate barely moves, from 6.5% to 8.3%.
Can I use this data?
Yes. The aggregate table (overall, by sector, trade, and metro, with counts and shares) is free under CC BY 4.0: the dataset page, CSV, JSON. Cite it as: CiteMy.ai (2026). Local business AI readiness: 1,238 local service business websites, 13 trades, 7 US metros, October 2026. https://citemy.ai/research/local-business-ai-readiness
Related research
- Phoenix contractor websites: AI readiness: the same five checks on Phoenix roofing, HVAC, and window contractors.
- AI search statistics: checked figures on how people are moving from search to AI assistants.
- Citation sources by engine: which websites ChatGPT, Claude, Perplexity, and Gemini cite for local questions.
Results describe what was measured, not a promise of ranking.