Sauce · AI and search
An AI assistant almost recommended you. Then your facts did not match.
Quick answer
When an AI assistant picks a business to recommend, it pulls your details from several places at once. Your website, your Google profile, directories, review sites. Then it checks whether the facts agree. If your address, hours, phone, or services do not match across those places, the assistant is not sure it has the right business. A business it is not sure about does not get named. The fix is dull but it works: one correct set of facts, the same everywhere.
Is your trade still open in your city?We take one client per trade, per city. The call starts with that check, not a pitch.
Book a free assessmentYou did the work. The AI still skipped you.
Your website is good. Your reviews are strong. A buyer asks ChatGPT or Google for "a good [your trade] near me". Your name is not in the answer.
Often the reason is not your website. It is that your business says three different things about itself in three places. The assistant could not tell they were all you.
How an AI assistant picks a business to name
It does not read one page. It gathers facts about you from many sources. Then it compares them and decides how sure it is.
Gather
The assistant pulls your name, address, phone, hours, and services from your site, your Google profile, map apps, and directory listings.
Cross-check
It lines those facts up. Same business, same details, in every source? Or do they disagree?
Score confidence
Matching facts raise its confidence that it found one real business. Facts that clash lower it.
Name or skip
When it is confident, it names you. When it is not, it plays safe and names a business it is sure about.
This is not new. Google's local ranking has used the same idea for years. Matching business data is a trust signal. AI just raises the cost of getting it wrong. The assistant gives one answer, not ten links, so being left off hurts more.
The five facts that have to match everywhere
Not your whole website. Just these five, word for word, in every place your business shows up.
- Name: the same name everywhere. Not "Delta Auto Care" on Google and "Delta Auto Care Ltd, Best Mechanic" on Yelp.
- Address: one format. "104 - 8829 Scott Rd" written the same way on every listing.
- Phone: one number, the one a person answers.
- Hours: the same on the website, the profile, and the door.
- Category and services: the same main trade, and the same service names, on the profile and the site.
- A name with extra keywords on one site but not another.
- "Suite 200" on Google, "#200" on Yelp, "Unit 200" on the website.
- An old phone number still live on a directory you forgot about.
- New hours on Google, old hours on the website.
- "Auto repair" on the profile, "full-service automotive centre" on the site, with no clear link between them.
Where your facts drift without you noticing
Go deeper
Related on the site
Each one takes a piece of this further. Pick what you need.
NAP consistency, explained
Name, address, phone: why the exact wording matters, and how to clean it up in order.
See the guide › Piece 02Local business entity SEO
Making the whole picture of your business consistent, so search engines and AI treat it as one thing.
See the hub › Piece 03Business info consistency for AI
The same idea, aimed specifically at how AI assistants read and trust your details.
See the guide ›See what the web says about your business.The free 49-point audit checks your details across the major directories and flags every place they disagree, plus whether an assistant can verify you at all.
Get the free scanThe 20-minute check you can do today
Search your name plus your city
Open every result on the first two pages. Google profile, Yelp, Facebook, Apple Maps, trade directories.
Write down five facts from each
Name, address, phone, hours, main category. One row per source.
Circle every difference
Every spelling. Every "Suite" versus "#". Every old number. Every wrong hour.
Fix the source, not just the copy
Correct it on the profile or directory itself. If a small directory gets its data from a provider, fix the provider so it stops re-spreading.
A smart assistant does not recommend a business it is not sure it has found.
What this is not
It is not about having more listings. A few listings that all agree beat dozens that argue.
It is not a one-time job. Details drift. A new phone system, a holiday hours change, a directory that resets. Check it twice a year.
It is not only an AI problem. Google has used matching data as a trust signal for years. Fixing it helps your map ranking and your AI visibility at once.
Open your own name in a search right now. If the first five results do not tell the exact same story about you, you have found this week's easiest fix.
Does mismatched information really make my business invisible to AI?
It does not delete you. It lowers the assistant's confidence that it has found one real business. When the assistant gives a single recommendation instead of a page of links, low confidence is enough to leave you off. It names the businesses it is sure about.
Which of the five facts matters most?
Name and address. They are how a machine tells one business from another. A wrong phone number or a wrong hour is a smaller problem. A name or address that reads differently across sources can look like two separate businesses.
How do I fix a listing on a site I do not control?
Most big directories have a "claim this business" or "suggest an edit" link. For the many small directories, the fix is usually upstream: correct the data providers that feed them, and the small listings update over time. A citation audit shows which ones are wrong.
Is an llms.txt file the answer to this?
No. Google says plainly that it does not read llms.txt and that publishing one does not change your visibility. What matters is that your real listings and your website agree. The lever is consistency across the sources an assistant checks, not an extra file.
How often should I re-check?
Twice a year for a stable business. And right after any change to your name, address, phone, hours, or a rebrand. A move or a rename is when the most damage happens, because the old version keeps living on listings you have forgotten.
Sources
- Google Search Central, “AI features and your website”: AI experiences in Search are grounded in Google's web results and follow the same content and quality guidance as the rest of Search.
- Google Business Profile Help, “Improve your local ranking on Google”: local results use relevance, distance and prominence, and complete, consistent business information helps Google understand and show your business.
- Google Search Central, “Consolidate duplicate URLs” and the local business structured data guidance: give search engines one clear, canonical version of an entity rather than several competing ones.
- Whitespark, Local Search Ranking Factors (2026 edition): citation consistency and NAP accuracy remain foundational trust signals for Local Pack visibility. Practitioner survey, directional.
Next step
Find out whether an assistant can verify your business
The free 49-point scan checks your details across the major directories and flags every conflict.