The llms.txt File That Could Decide Whether AI Search Ever Reads Your Plumbing or Electrical Site
A new plain text file is quietly deciding which pages AI assistants actually read on a plumbing or electrical website. Here is what llms.txt does, and why almost no trades business has published one yet.
A boiler stops heating at half past six on a Sunday and the homeowner does not scroll through a directory. They open ChatGPT, describe the problem, and ask for a plumber or heating engineer nearby who can come out today. Before the model names anyone, something has already decided which of the dozens of trades sites in that town are even worth reading, and for almost every plumbing and electrical business, that something has never heard of their site at all.
What llms.txt Actually Is
llms.txt is a plain text file published at yourdomain.com slash llms.txt. It was proposed in 2024 as a short, curated summary a site hands directly to an AI model, written in sentences a model can use immediately instead of guessing from a navigation menu, a cookie banner, and a stock photo carousel. A growing number of AI tools check for this file before they read anything else on a site.
1. A Trades Site Has A Dozen Pages And Most Models Only Read Two
A typical plumbing or electrical website has a homepage, a services page, an about page, a contact page, and maybe a handful of town specific pages if the business has invested in those already. That sounds small compared to a law firm's site, but a model answering an urgent question still only has a few seconds to decide what to read before it writes an answer. Left on its own it usually stops at the homepage. An llms.txt file tells it exactly which of those pages actually answers the question someone is asking right now.
2. What Belongs In The File For A Trades Business
- A one paragraph description of what you actually do and the towns you cover, not a slogan
- A direct link to your callout fee and pricing page, since that is the question most urgent searches are really asking
- Your trade license or registration number and the body that issued it, stated as text in the file itself
- A link to the specific page for emergency or same day availability, separate from your general services page
- A link to your three or four strongest town specific pages, not a single shared areas we cover page
3. The File Points. It Does Not Replace The Proof
An llms.txt file only earns a model's attention for a few extra seconds. If it points to a page that still says call for a quote with no callout fee, no license number, and no stated response time, the extra attention does nothing, because there is still nothing on the page worth repeating. ZENETERA has covered what belongs on those pages in earlier posts on callout fees, credential numbers, and service area pages. The file gets a model there faster. It does not fix a page that has nothing to say once it arrives.
| Without an llms.txt file | With one that points to the right pages |
|---|---|
| A model reads whatever page it reaches first, usually the homepage, and stops there | A model is pointed straight at your pricing, availability, and credential pages in one pass |
| An old blog post competes equally with your callout fee page for attention | Your most useful pages are named directly as the ones that matter |
| Every plumbing or electrical site looks roughly the same to a model skimming quickly | A business that states exactly what it wants read stands out by being easier to use |
An llms.txt file does not make a vague page specific. It makes sure a specific page actually gets read before the model moves on.
The Takeaway
Almost no plumbing or electrical business has published an llms.txt file yet, mostly because almost none have heard the term. The ones that add one this year, pointing straight at a callout fee page, a credential number, and a handful of town specific pages that already say something concrete, give AI assistants a faster, clearer path to the exact proof that gets a business named instead of skipped. That small head start matters most on the jobs that get booked within minutes of being asked.