Keeping Business Profile, Bing and your site in agreement
Google Business Profile, Bing Places, directories and your own site each hold a copy of the same facts, and the copies drift. The weekly diff shows where, and the fix is one source feeding all of them.
Quick answer: Assistants read a business's name, address, hours and services from the Business Profile, Bing, directories and the site, and the copies drift; a weekly diff against the site's structured data finds each difference and one CMS record should feed all surfaces.
- AI Search Visibility
- Measurement
- Machine-Readable Delivery
Quick answer
Why do AI assistants give a different phone number or opening hours from your website?
Because they read several surfaces and pick one. Google Business Profile, Bing Places, directories and your own site each carry a copy of your name, address, hours, categories and services, and the copies drift. A weekly diff shows where, and the fix is to make them agree at the source.
Additional Context
Microsoft Copilot draws its web results from Bing's index, so a site that has never verified Bing Webmaster Tools is relying on a surface it has never looked at.
Sources
- Google Business Profile Help - Guidelines for representing your business
Google's rules for the name, address, hours, categories and other details on a Business Profile.
- Bing Webmaster Guidelines
Bing's published guidance on how it discovers, crawls, indexes and ranks content.
- Schema.org - LocalBusiness
The vocabulary for describing a local business, including name, address, telephone, opening hours and identifiers.
The entity record
Copilot reads Bing
Many businesses verify Google Search Console and never open Bing Webmaster Tools. Microsoft Copilot draws its web results from Bing's index, and Bing Places holds the listing it shows for a business. A business that has never claimed either is being described from whatever Bing found on its own. Verifying both is a short job and adds a second index to the weekly record, with its own crawl errors and its own view of which pages exist.
The entity record
The facts an assistant repeats about a business are a short list. Name, address, phone, opening hours, the categories the business belongs to, the services it offers, and the licence or registration numbers that apply to it. Each one lives in several places: the Business Profile, Bing Places, the directories that scraped the profile years ago, the footer of the site, the contact page, and the structured data in the page head. Google's guidelines for a Business Profile set out what the name and categories may contain, and the Organization and LocalBusiness vocabularies on schema.org give the site a way to state the same facts in a form a machine reads directly.
The site's structured data is the business's own statement, and it is the one copy the business fully controls. When it is generated from the same record in the CMS that renders the contact page, the two cannot disagree. When the Business Profile is updated from that same record through the API, there are no longer three copies to keep in step, only one record and three outputs of it.
What drift looks like
The business moved and the old address is still on two directories. Holiday hours were set on the Business Profile and never on the site. A service was added to the profile that the site has never mentioned, so the assistant finds a claim with no page behind it. The phone number on the Bing listing is a former staff member's mobile. None of these is dramatic on its own. Together they are why an assistant hedges, gives two answers, or gives the wrong one, and why a customer who asked an assistant turns up at the old address.
Keeping them in agreement
The weekly record pulls each surface and diffs it against the site's own structured data. A difference is a finding with the surface, the field and both values recorded. The engineering response is to make the site's record the source, generate the structured data from it, push the Business Profile from it, and leave the directories a dated list of corrections to request. This is the entity dimension of AI Visibility Management, and it is a common cause of a wrong answer found in the assistant answer record. How the site states the facts for machines in the first place is covered under machine-readable delivery.
Key Takeaways
What keeps a business's facts consistent across surfaces
- Copilot draws its web results from Bing, which many businesses have never verifiedImportant
Claiming Bing Webmaster Tools and Bing Places adds a second index to the record, with its own crawl errors and its own view of the site, and it is a short job.
- The entity facts are a short list held in many placesCritical
Name, address, phone, hours, categories, services and registration numbers each live on the profile, the listings, the directories and the site, and the copies drift.
- The site's structured data is the one copy the business fully controlsCritical
Generated from the same CMS record as the contact page, it cannot disagree with the page, and it is the statement an assistant can read without guessing.
- One record and three outputs beats three copies kept in stepImportant
Push the Business Profile from the site's record through the API and leave the directories a dated correction list.
Assistants repeat a short list of entity facts read from several surfaces, and the copies drift. A weekly diff against the site's own structured data finds each difference, and the engineering fix is one source in the CMS feeding the page, the structured data and the Business Profile.
Questions about listing consistency
Do we really need Bing Webmaster Tools?
If any customer uses Copilot, Edge or Windows search, yes. Bing's index is where their web results come from, and verifying the tool shows what Bing has crawled and what it could not reach. It costs nothing and it does not change the site.
Which copy of our details is the right one?
The site's own, published in HTML and in structured data from one record in the CMS. Every other surface is an output of it or a correction to be requested.
Can the Business Profile be updated from our site automatically?
Yes, through the Business Profile API, once the business's Google Cloud project has been approved for it. Until then the weekly diff lists the fields to change by hand.
What about directories we never signed up to?
They scraped a profile at some point and will not update on their own. The record keeps a dated list of each one, the wrong value and the correction requested, so the business can see which ones have been fixed and which still contradict it.
