Knowing how to check if ChatGPT recommends your business starts with one shift: stop asking once and start counting. Pew Research Center reports that the share of US adults who say they use ChatGPT rose from 18% in 2023 to 44% in 2026. Marketing leads at property management and real estate firms are right to wonder whether their company comes up when a prospect asks an assistant for a recommendation.
This guide covers a repeatable way to check, where AI visibility numbers come from, what they cannot show, and how to report a trend over months instead of a one-off score. Its focus is measurement. For the wider picture, see our guide on how to use AI for property management.
Ask AI assistants the questions your prospects ask, leave your company name out, ask each question several times and count how often you are named. A single answer on a single day tells you almost nothing.
AI assistants such as ChatGPT, built by OpenAI, write a fresh answer every time. One reading can catch a good day or a bad one. A count across many attempts is something you can compare next month.
The check has four steps:
Start with a short, fixed list of questions written the way a buyer types them, so every month measures the same thing.
Leave your company name out. Asking whether your firm is any good only shows that the assistant can read your name back to you. Prospects ask comparison and shortlist questions instead, such as “Which CRM implementation consultants specialize in configuring HubSpot for property managers?” or “Best CRM implementation services for property management firms.” Because our founders came from property management, we write these the way operators actually ask.
Save each prompt’s exact wording, the assistant you used and the date.
Asking each question several times turns a yes-or-no into a count. Use a fresh or temporary chat, signed out where the assistant allows it, so past conversations do not skew the answer.
Repeat the set in ChatGPT, Perplexity, Gemini and Google AI Mode. Record the result as a count, for example “named in 3 of 5 attempts,” rather than yes or no. That figure can move, and movement is what you are watching for.
Also note who is named instead. Those names show who the assistant currently treats as the answer.
Being named and being cited are two different results. Named means the assistant mentions your company in its answer. Cited means it links to your website as a source.
Log them in separate columns. Then check which websites the answer drew from, such as review sites, directories or industry publications. Those sources show where coverage matters for the questions your buyers ask.
What to do with that list is a separate job: answer engine optimization (AEO), the work of making your content easy for AI tools to find and quote. Our guide to AEO for property management and real estate covers it.
AI visibility data comes from three places: Google Search Console’s generative AI report, tools that repeat a fixed set of prompts, and checks you do by hand. Each one sees a different slice, and none of them sees everything.
Google launched the Search Console generative AI performance report on June 3, 2026, and rolled it out to all sites on August 31, 2026. It shows impressions, meaning the times a page appeared, in Google AI Overviews and Google AI Mode, by page, country, device and date. It covers Google only.
Prompt trackers, such as HubSpot’s AEO tool, ask the same questions again on a schedule. We track our own buyer prompts this way. Hand checks are useful too, but they are a sample and should be labeled as one.
| Source | What it shows | What it misses | Real or sample |
|---|---|---|---|
| Search Console generative AI report | AI Overview and AI Mode impressions by page, country, device and date | Clicks, search queries, assistants outside Google | Real data, Google only |
| Prompt tracker | Who is named for a fixed set of prompts, repeated on a schedule | Questions outside your set | Repeated sample |
| Hand checks | Full answers to your own prompts | Scale and consistency | Small sample |
Whatever the source, every number should say where it came from and whether it is real or sample data. That habit is what makes any of the key metrics property managers track worth reading.
They cannot tell you how many visits or leads came from AI answers, because Google counts clicks from AI Overviews as part of normal organic search traffic. They also cannot give you a fixed rank, because the same question can get a different answer each time.
Google’s report has no clicks, no click rate and no search queries, only impressions. In Search Console and Google Analytics 4 (GA4), clicks from AI Overviews sit inside ordinary organic data and cannot be separated out.
Answers also change from one attempt to the next and from person to person, so there is no fixed position to chase.
If your company is not mentioned yet, that is common. Treat it as the starting point of your log.
Report AI visibility as a trend over months: ask the same prompts the same way on a set schedule and log the results. A line that moves month by month is an honest measure. A single screenshot can mislead.
A simple monthly log for AI visibility tracking needs four columns:
Keep Search Console AI impressions beside the log as a separate measure, because the two count different things. Where a figure is missing, write “Not available” and the reason, never zero. A zero says you checked and found nothing.
Read the trend over several months. No single month is the story. Revisit it the way you would revisit a SWOT analysis for property management rather than file it once.
Geekly Media was founded by former property managers and builds reporting inside HubSpot, so AI visibility can sit next to the rest of your marketing numbers rather than in a separate spreadsheet.
As a HubSpot Diamond Solutions Partner, we build dashboards that put a prompt log beside your property management marketing reporting. On the strategy side, SEO support, including property management SEO content, and fractional CMO work help decide which questions to track. We do not promise mentions, citations or leads. We help you measure them the same way every month.
Speak to a Geek
No. ChatGPT’s answers change between attempts and between users. A Washington State University study published in March 2026 found that across 10 identical prompts, ChatGPT gave consistent answers on only 73% of statements. That is why a count across several attempts tells you more than any single check.
Not separately. Google Analytics 4 counts clicks from AI Overviews as ordinary organic search traffic, and Search Console’s generative AI report shows impressions only. Report AI impressions and mentions as their own measures, and avoid labeling any figure as AI traffic, because that number cannot be pulled out.
Monthly, using the same prompts and the same method each time. Daily checks mostly pick up random variation from one attempt to the next rather than real change. Review the trend over a quarter or longer before drawing conclusions, and keep the log consistent so each month compares fairly with the last.