What We Rebuilt So 33 Websites Could Talk to AI Search
We spent last week doing something unglamorous across every website we host, and the result is worth explaining because most business sites have the same problem. Search engines and AI assistants build a picture of your business from structured data on your site — a code block stating your name, address, phone, hours and services. Get it right and you’re quotable. Get it wrong and you’re a guess. Here’s what we actually found when we looked. Twenty sites were publishing two competing business listings. Four were publishing three or more. This happens because SEO plugins each add their own, and a site that has picked up two or three plugins over the years ends up describing itself two or three times, often with different details. On two sites the competing listings carried different business names. A search engine reading that has to choose which version of you to believe. It usually chooses badly. So we went through all thirty-three. One listing per site. Real street address or, for the businesses that work from a van rather than a shopfront, a proper list of the towns they serve. Phone numbers, opening hours, services in the words a customer would actually use. For a client with two branches, both locations published properly and linked, so nobody thinks one building moved. We finished with every site publishing exactly one clear listing — and a dated record of every check on every site, so the work is provable rather than claimed. If you don’t know whether your site does this correctly, that’s normal, and it’s not something you can see by looking at the page. We’re happy to check.
Is Your Website Invisible to ChatGPT and AI Search?
More and more people are asking an assistant instead of typing into a search box. “Who does vehicle wraps near Tupelo?” “Is there a place in Iuka open late that sells wine?” Whatever answer comes back was read off somebody’s website. If the crawlers those assistants use can’t read yours, you’re not in the running. Not ranked low — absent. We went through every site we host last week looking for exactly this, and we found sites quietly refusing the crawlers that feed ChatGPT, Claude, Perplexity and Google’s AI answers. Nobody chose that. It came in with a theme, or a plugin, or a file somebody pasted in years ago and forgot. Here’s how to check your own. Type your domain into a browser followed by /robots.txt — so yourbusiness.com/robots.txt. You’ll get a plain text file. Look for any line that says Disallow: / underneath a name like GPTBot, ClaudeBot, PerplexityBot, Google-Extended or CCBot. If you see one, that crawler has been told to go away. Two things worth knowing. A file that doesn’t exist at all is usually fine — it means nothing is being blocked. And a file that’s there but blocking ad networks and scrapers is often deliberate and harmless. It’s the AI and search crawlers you care about. The other trap is harder to spot: a physical robots.txt file sitting on the server overrides whatever your SEO plugin thinks it’s publishing. Your plugin shows one thing in the dashboard, visitors get another. We found that on twenty-six sites. If you’re not sure what you’re looking at, send us the address and we’ll look for free. It takes a minute and it’s the kind of thing that costs you customers silently for years.