llms.txt is the latest SEO buzzword, but does it actually work? No. There is no evidence that an llms.txt file improves rankings, AI citations, or lead flow. Over a 31-day monitoring period across local service business websites, Googlebot made zero requests for the file. Your SEO budget belongs on local-intent content, entity signals, and speed to lead, not on a text file that no search engine uses.
If you run a local service business, you have probably heard the pitch by now. Some agency or "AI SEO expert" tells you that the future of search depends on a single file called llms.txt. Add it to your site, they say, and AI models will suddenly know who you are, what you do, and why they should recommend you. It sounds technical, important, and exactly like the kind of thing you might be missing.
Here is the reality. llms.txt may help organize your site for humans and agents, but current evidence does not show that it improves rankings, citations, or lead flow (SE Ranking, 2025). Before you pay anyone to set it up, here are the six myths costing contractors real money.
Myth 1: "llms.txt will get your site cited by AI"
The biggest promise you will hear is that this file acts like a beacon, telling AI models to cite your business as an authority. In reality, no major AI vendor has confirmed that publishing an llms.txt file increases citations. Recent industry guidance confirms that it has no effect on Google Search or AI Overviews (Search Engine Land, 2026).
AI models prioritize information based on entity signals, user trust, and content relevance. Not on a single text file sitting in your root directory. Think of it this way: if a homeowner asks ChatGPT for a roofer in your city, the AI is looking at what the rest of the web says about you, not what you say about yourself in a hidden file. For a deeper look at how AI search actually picks winners, check out Is ChatGPT Recommending Your Business? How AI Search Actually Picks Winners in 2026.
Myth 2: "llms.txt is the new must-have SEO file"
Agencies love to pitch this as the next robots.txt or sitemap.xml. It is not. It is currently a proposed convention, not an official standard. Most current writeups describe it as a form of content hygiene, which is a way to tell AI models what parts of your site are safe to read, rather than a ranking mechanism (SE Ranking, 2025).
Treating it as a "must-have" is a classic example of technical theater that distracts from actual growth work. If your service pages are thin, your Google Business Profile is incomplete, and your reviews are sparse, no text file is going to rescue your visibility. The 8 things AI search engines look for before they recommend a local business have nothing to do with llms.txt and everything to do with the fundamentals.
Myth 3: "It works because AI crawlers use it today"
The claim that "everybody is doing it" does not mean it actually does anything for your bottom line. Over a 31-day monitoring period across local service business websites, Googlebot made zero requests for llms.txt. If the biggest search engine in the world is not even looking for the file, it is certainly not using it to decide whether your HVAC or roofing company deserves to appear in a search result.
This is the part where the "experts" usually pivot. They will say something like, "Well, it is not about Google, it is about the AI crawlers." But when pressed for evidence that any specific AI crawler reads or uses the file to influence recommendations, the evidence is thin to nonexistent. The file costs nothing to create, which is exactly why agencies can charge for it with a straight face. You are paying for the appearance of technical sophistication, not for results.
Myth 4: "Technical setup beats content"
There is a dangerous tendency to believe that a few lines of code can replace the hard work of building a great website. Search coverage still depends on strong pages, clear service-area content, and entity signals. If your service pages are thin or your local citations are inconsistent, no technical file will save your rankings.
Your SEO budget is better spent on answering the questions your customers are actually typing into search bars than on experimental files that do not move the needle. A homeowner searching "why is my AC blowing warm air" is looking for an answer and a local pro who can fix it. That is the content that earns clicks, calls, and citations from AI engines. Not a file that tells crawlers where your markdown lives.
Myth 5: "AI SEO experts can guarantee leads from AI search"
Any agency promising a "guaranteed" increase in leads by adding an llms.txt file is selling a fantasy. No current source supports a guaranteed lead impact from this file. When a salesperson uses terms like "AI-powered" or "guaranteed AI visibility" to sell you a simple text file, they are usually inflating the value of a low-effort task to justify a high-effort price tag.
Here is a simple test. Ask the agency what specific data they have that links llms.txt to measurable lead growth for local service businesses. If they cannot point to a single study, a single case, or a single piece of evidence, you have your answer. The file takes only minutes to generate. If someone is charging hundreds or thousands for it, you are not buying results. You are buying theater.
Myth 6: "If competitors do it, you have to"
Current coverage still treats llms.txt as experimental and unproven (SE Ranking, 2025). Just because a competitor has a file on their site does not mean they have discovered a secret ranking factor. They likely hired an agency that sold them on the same myth.
Focus on what actually brings in calls: being the business that shows up when a homeowner has a problem and needs a pro. The contractors winning right now are the ones investing in GEO for contractors: getting cited when homeowners ask AI who to hire, not chasing experimental files that no search engine reads.
What Actually Gets Local Contractors Cited by AI
Here is where the conversation gets interesting. The contractors getting cited by AI search engines are not the ones with the most technical files on their servers. They are the ones with detailed service pages that answer specific local questions, complete Google Business Profiles with recent reviews, and consistent business data across every directory.
Worked example (illustration, not a client result): Consider two HVAC contractors in the same metro area. Contractor A runs a 15-page site with detailed service-area pages answering questions like "why is my AC blowing warm air," 40-plus Google reviews mentioning specific repairs, and consistent name, address, and phone number across every directory. Contractor B has a basic 5-page site, 12 reviews, and an llms.txt file. When a homeowner asks an AI engine for an HVAC recommendation, Contractor A gets cited because the AI finds corroborating signals across multiple sources. Contractor B's llms.txt file adds nothing the AI can use.
When a homeowner asks ChatGPT, Perplexity, or Google AI Overviews for a recommendation, the AI is looking for corroborating signals. It wants to see the same business name, address, and phone number across multiple sources. It wants to see reviews that mention specific services like "AC repair" or "roof replacement." It wants to see service pages that match the intent behind the question. No llms.txt file required.
The shops that invest in these fundamentals are the ones showing up in AI answers. The shops spending their budget on experimental files are not. That is the pattern, and it lines up with current industry guidance (Search Engine Land, 2026).
The Reality: Where Your SEO Budget Should Go
If you want to win in 2026, stop looking for magic files and start looking at your foundation. Visibility in AI answers is not controlled by llms.txt based on current published guidance (Search Engine Land, 2026).
Instead, focus on these three areas:
- Local-Intent Content. Write pages that answer the specific questions your customers ask, like "how much does a new roof cost in Charlotte" or "why is my AC blowing warm air." These pages are what AI engines lift and cite when homeowners ask for help.
- Core Entity Signals. Ensure your name, address, and phone number are consistent across every platform. Your Google Business Profile, your website footer, your Yelp listing, and every directory in between should all tell the same story.
- Speed to Lead. Older research, including the MIT-affiliated Lead Response Management Study (2007), suggests that contacting a lead within five minutes dramatically increased conversion odds compared to waiting even ten minutes. Recent field-service platform data continues to reinforce that speed-to-lead remains a critical conversion factor (Housecall Pro, 2025). For the full breakdown, read The 5-Minute Rule: Why Speed-to-Lead Beats Marketing Spend.
If you want to make sure your site is ready for the future of search, read the guide on Is ChatGPT Recommending Your Business? How AI Search Actually Picks Winners in 2026. And do not forget the back door most local businesses skip: Bing Is the Back Door to ChatGPT: The Setup Most Local Businesses Skip.
