A duplicate listing splits one business into two: your reviews, photos, and profile views divide across the pair, the facts on the two profiles drift apart, and a machine reading both has no way to know which version of you is real. For an AI assistant deciding whether to name your business — a decision that runs on confidence and consistency — a split identity is close to the worst input possible. The good news is that this is one of the most findable, most fixable problems in AI visibility, and the fix has a documented track record. Here’s how the damage works, how duplicates happen, how to find yours, and how to resolve them.
What a split identity does to a machine
Assistants judge truth by comparison. They have no memory of your business — they retrieve sources in the moment and cross-check them against each other. Two profiles for one business sabotage that process twice over:
The signals divide. Reviews land on whichever listing a customer happened to find. Photos, questions, and engagement split the same way. Neither profile ends up looking like the established, well-documented business you actually are — you’ve turned one strong candidate into two weak ones.
The facts contradict. Two profiles almost never say identical things. One holds the old address, the old hours, a dead phone line; one holds the new. A machine comparing them sees a contradiction it cannot resolve — is this one business that moved, or two businesses? Which phone rings? Confidence is the quiet gatekeeper of AI answers, and the safest move for an uncertain assistant is the one that hurts most: leave you out and name a competitor whose story checks out.
Notice that neither failure requires either listing to be mostly wrong. The mere existence of the pair is the problem — division plus contradiction, regardless of how the details shake out.
How duplicates happen
Almost never on purpose. The common paths:
- A move. You relocate, create or update a profile at the new address, and the old-address listing lives on — sometimes maintained by nobody, still confirmable by old citations and stale directories all over the web.
- A reboot. A rebrand, an ownership change, a recovered-then-recreated account. Someone sets up a fresh profile instead of updating the existing one, and now both exist.
- Third-party imports. Directories, data aggregators, and booking or delivery platforms generate business listings at scale from public records — and those records can spawn an unclaimed profile you never knew existed.
- Helpful hands. A past employee, a marketing vendor, even Google’s own suggestion pipeline creating a listing that didn’t match the one already there — slightly different name, suite number formatted differently — so nothing flagged it as the same place.
Because none of these announce themselves, duplicates are typically discovered the way most source problems are: someone finally goes looking.
What it looked like in practice
This isn’t hypothetical for us — a duplicate profile is at the center of the most documented engagement we’ve published.
In November 2025, a Sacramento window-tinting shop — over a decade in business, 5.0-star reviews — was fading on Google: its monthly Business Profile report showed profile views down sharply, with zero calls for the month. The tell came by email: Google sent two separate November reports for the same business. The shop had two profiles. One was the listing customers actually used. The other drew 58 profile views that month and produced nothing — no calls, no website visits, not a single interaction. Every one of those views was a potential customer looking at the wrong version of the business.
The fix was the unglamorous playbook: consolidate down to one listing, then build out the surviving profile — services, service area, hours, photos — with the details kept consistent everywhere they appear. Over the following months the shop’s own Google reports recorded the turnaround: profile views recovering and then climbing 79% in a single month by March 2026, calls returning, total interactions three times the November count. By late June 2026, Google’s AI Overview answered “mobile window tinting in sacramento” by naming this shop first.
Every claim in that sequence traces to dated documentation — Google’s own report emails, dated search captures — and the honest limits are stated alongside it: two months of reports are missing from the record, and a documented sequence is not a guaranteed outcome. The full case study, with the recreated evidence panels, is on our results page.
How to find your duplicates
Fifteen minutes, no tools:
- Search Google Maps for your business name — then for common variants: with and without “LLC”, the old name if you rebranded, likely misspellings.
- Search every address you’ve ever operated from. Old-address listings are the classic survivor.
- Search your phone numbers, including any old lines and tracking numbers a vendor may have published.
- Ask the assistants. Run the factual questions from our DIY check — “what are [your business]’s hours and phone number?” — and watch for answers that mix two versions of you, hedge between addresses, or describe the business inconsistently across runs. A split identity often shows up in the answers before you’ve found the second listing itself.
Anything you find that represents your business and isn’t the profile you maintain goes on the fix list — including unclaimed stubs with no reviews. Machines read those too.
The fix path
The sequence matters: claim, then merge or remove, then clean up what fed it.
- Claim what you can. You can’t fix a listing you don’t control. Google’s process for claiming and managing listings — and for reporting duplicates of a verified profile — is documented at support.google.com/business.
- Merge or remove the extras. For true duplicates, Google can merge or remove the redundant listing; a moved-location listing gets marked accordingly rather than left contradicting the new one. Reviews on a merged duplicate are the detail people worry about most — handle this through Google’s own process rather than deleting anything hastily.
- Fix the sources that spawned it. If an aggregator or directory holds the old address, it can regenerate the problem or keep corroborating the wrong version. Correct the record where the bad data lives, or the contradiction outlives the duplicate.
- Re-check. A few weeks later, re-run the searches and the assistant questions. Source fixes take time to propagate into answers — what you’re looking for is one profile, one address, one consistent story everywhere.
Why we fix this first
When we audit a business, entity problems — duplicates chief among them — are the gap we clear before any other, because everything else builds on a single, unambiguous identity. Schema markup restates your facts machine-readably; a well-maintained Business Profile feeds assistants those facts constantly. Neither helps while there are two of you, because a contradiction stated more clearly is still a contradiction.
If you’d rather have the search done for you, the free check includes it: what three assistants currently say about your business, which sources they cited, and the first gap we’d fix — which, more often than owners expect, turns out to be a second version of themselves they didn’t know was out there.