How to check whether ChatGPT, Claude and Gemini know your CRE firm
Tenants, investors and owners increasingly ask an AI assistant for a shortlist before they search. If your firm is not in that answer, you never learn you were passed over.
Why the three engines disagree
The surprise for most firms is that the engines do not agree with each other. For the same firm, one gives an accurate summary, another has an old office address and a partner who left two years ago, and a third has never heard of them and recommends competitors.
Each engine trains on and retrieves from different sources, so checking one and assuming the rest match is a mistake. That is why the test runs all three.
What the test actually asks
- Category questions — "who are the best [asset type] brokers in [market]?" This is the question that produces a shortlist, and the one you want to appear in.
- Entity questions — "what do you know about [firm]?" This surfaces wrong addresses, outdated teams and the wrong specialty.
- Competitor questions — which firms the engine names instead of you, in its own words.
What usually keeps a CRE firm out of the answer
- No page answers the question. A single "Services" page cannot be quoted for "industrial tenant rep in the Southeast Valley".
- Firm facts differ across sources — a different address, phone number or description on your site, LinkedIn and directory listings. Inconsistency makes a model hedge.
- No citable third-party mentions — trade publications, associations, local business journals.
- Broker profiles are thin, so there is nothing to attribute expertise to.
Checking it yourself vs. tracking it
You can run these prompts by hand in three browser tabs. The two things that are hard by hand are consistency (the same prompts, worded the same way, every month) and attribution (which engine changed, and when). The AI Visibility & Citation Tracker records mentions and citations over time and benchmarks them against named competitors.
In DeshTEK Digital
- Runs against Claude, ChatGPT and Gemini in real time
- Costs 2 credits per run; the free plan includes 5 credits a month
- Paired with the AI Visibility & Citation Tracker (8 credits per prompt) for tracking over time
Run it on your own site
The free plan includes 5 credits a month — no credit card, no time limit.
Start free →Frequently asked questions
How do I check if ChatGPT knows my commercial real estate firm?
Ask ChatGPT "what do you know about [your firm name]?" and "who are the best [asset type] brokers in [your market]?", then compare the answer with what your website claims. DeshTEK Digital's AI Search Visibility Tester runs both question types against ChatGPT, Claude and Gemini at once and reports what each says.
Why does ChatGPT recommend my competitor instead of my firm?
Usually because the competitor has pages that answer specific questions, consistent firm details across the web, and third-party mentions the model can cite. Deal volume does not help if none of it is published in a form a model can read.
Do Claude, ChatGPT and Gemini give the same answer about a firm?
Often not. They train on and retrieve from different sources, so one engine can describe a firm accurately while another has outdated details or has never heard of it. Check all three.
How often should a CRE firm check its AI search visibility?
Monthly is enough for most firms. Check again after a website rebuild, a rebrand, an office move or a merger, since those are the changes that leave stale facts in AI answers.
Is AI search visibility the same as SEO?
No. SEO earns a click from a list of links. AI search visibility, sometimes called GEO or generative engine optimization, is about being named and cited inside an AI-written answer. Strong SEO basics help, but the work of publishing direct answers and consistent firm facts is specific.