By Sumit Nautiyal, Founder · 6 min read · updated 6 August 2026
Why Bing Places decides your ChatGPT visibility
Short answer
ChatGPT and Copilot both resolve local queries largely through Bing's index, so an unclaimed Bing Places listing makes a practice close to invisible on two of the four major engines — regardless of how strong its Google presence is. Claiming it takes about an hour per location and costs nothing.
The finding we see most often
Run a practice through all four engines and a particular shape shows up again and again: strong on Gemini, almost nothing from ChatGPT and Copilot.
It looks like a content problem. It usually is not. It is an unclaimed Bing Places listing.
ChatGPT and Copilot both lean heavily on Bing’s index for local resolution. Gemini and AI Overviews lean on Google. If your Google Business Profile is immaculate and your Bing listing does not exist, you have optimised for one half of the market and disappeared from the other.
Why nobody does it
Bing’s consumer search share is small enough that most marketing teams stopped thinking about it a decade ago. That reasoning was defensible until roughly 2024. It is not now, because the traffic no longer arrives through bing.com — it arrives through an assistant that happens to read Bing underneath.
The practical result is a genuinely uncontested surface. In competitive metros where every practice has fought over Google for years, Bing listings are frequently unclaimed, incomplete, or carrying an address from two moves ago.
That asymmetry will not last. It is worth using while it does.
What to check, in five minutes
- Search your practice name on bing.com. Is there a complete business panel — hours, services, photos, phone?
- Go to bingplaces.com and search for your business. Is it claimed, or is it an unverified auto-generated record?
- If you have several locations, check each one. Partial claims are common — one flagship location verified, the rest ignored.
An unverified auto-record is worse than nothing, because it usually carries stale data that Bing trusts more than your website.
What to fix, in order
Claim every location. Not just the flagship. Engines resolve locations independently, so five locations with one claimed listing means four invisible practices.
Match your NAP exactly. Name, address and phone should be character-identical to your Google Business Profile. “Suite 200” in one place and “Ste 200” in another is enough to stop an engine connecting the two records to the same entity.
Fill every field. Hours, categories, services, description, photos. Sparse listings get cited less than complete ones, and the fields are free.
Add the services you actually want to be found for. If Botox is 40% of revenue, it should be named in the listing, not implied by the category.
The equivalent work on the Google side — which decides the other two engines — is in how to optimise Google Business Profile for AI search.
Foursquare, while you are there
Foursquare remains a significant structured local-data provider that feeds several assistants, and almost no healthcare operator thinks about it at all. Search your practice at foursquare.com. Stale addresses and closed-location records propagate further than people expect, and the fix is the same shape: claim, correct, complete.
What this will not fix
Claiming a listing does not make an engine recommend you. It makes you eligible to be recommended.
Once the listing exists and is accurate, the ranking work is elsewhere: review volume and recency, third-party editorial coverage, community presence. A complete Bing listing with eleven reviews will still lose to a competitor with two hundred.
Think of it as necessary and unglamorous groundwork — the cheapest item on the list, and the one that blocks everything downstream if skipped.
How to know whether it worked
Re-run the same prompts across ChatGPT and Copilot in a fresh, signed-out session about three to four weeks after claiming. Directory changes propagate faster than review or editorial work, so this is one of the few interventions where you can reasonably expect visible movement inside a month.
Measure before you change anything, or you will not be able to tell.
The usual caveat
No AI platform publishes how it selects sources. Everything here is drawn from measured output and correlations documented across the industry — not confirmed ranking factors. What is certain is narrower and still useful: ChatGPT and Copilot resolve local queries through Bing, and you cannot be selected from an index you are not in.