VisibleAI

See if ChatGPT, Gemini & Perplexity recommend your business — or your competitor. Free AI visibility scan, real answers, clear fixes. Data posts weekly.

[VisibleAI @usevisibleai]

We asked ChatGPT to recommend a CRM 165 times.

5 names — Zoho, HubSpot, Pipedrive, Salesforce, Freshworks — showed up in 70–84% of answers.

It suggested just using Gmail (41%) more than it named most CRMs.

SalesNexus: named 0 times.

[VisibleAI @usevisibleai]

Full study: https://t.co/SBc5uM2jc8

[VisibleAI @usevisibleai]

We asked ChatGPT to recommend project management software 210 times.

Asana: named in 91% of answers.
ClickUp and Monday: 86% each.
One of those three appears in 94% of all answers.

20 established PM tools were named zero times.

Is your category this concentrated?

[VisibleAI @usevisibleai]

@aleyda If fewer visits reach the open web, more of the shortlisting is likely happening before any click, inside the answer itself. Do you think brands should start reporting how often AI assistants name them next to traffic, or is that number still too noisy to trust?

[VisibleAI @usevisibleai]

@harleyf Seeing how buyers find you will matter more than any one snapshot. When we re-ran the same buyer questions in ChatGPT, 6 of 8 brands had a likely score range 28 to 30 points wide. Will merchants see a trend over time, or one rolling 28-day total?

[VisibleAI @usevisibleai]

@OpenAI Will the insights separate being mentioned from being recommended? In our ChatGPT scans a CRM named in all 15 buyer answers was the actual recommendation in 67% of them, so the two numbers can tell quite different stories.

[VisibleAI @usevisibleai]

@Kevin_Indig Leaving out name-only mentions is a big gap for buyer research. In our ChatGPT scans the email marketing leader was named in 15 of 15 answers, and that would only show in the hub if the answers pointed to its site. Did the top 5 discovery topics match what your readers ask?

[VisibleAI @usevisibleai]

@theinformation If AI Mode carries fewer ads, more of the buyer's shortlist gets decided in the unpaid answer. In our ChatGPT scans, a smaller email tool made it into just 3 of 15 answers to the same buyer questions. Do you expect Google to put ads inside the answer or keep them around it?

[VisibleAI @usevisibleai]

@Techmeme @jaypeters The claim-your-profile part is what marketers will feel first. A profile tells ChatGPT who you are, but in our CRM scans the category leader was recommended in 60% of the answers that named it. Will the hub show which questions you get recommended for, or just mentions?

[VisibleAI @usevisibleai]

@Marie_Haynes Better intent reading may matter most on "which tool should I use" questions, where the model decides which brands to put forward, not just which pages to cite. Are you planning to run the same prompts on Ultra and standard AI Mode to see if the picks shift?

[VisibleAI @usevisibleai]

@chris_nectiv Mentions will be the number everyone checks first, but in our scans a mention isn't a pick: an email marketing leader was named in 15 of 15 ChatGPT answers and recommended in 11. Any sign the Hub separates the two, or does every mention count the same?

[VisibleAI @usevisibleai]

@glenngabe @juokaz Filling the wait while an image renders is a clever slot commercially. The bigger test is whether people can still tell the paid placement from the answer once ads show up next to recommendations. Do you expect labelling as clear as Google's?

[VisibleAI @usevisibleai]

@CrystalontheWeb Desktop is also where a lot of B2B software shortlisting happens, so for SaaS brands that 15% figure may be the one to watch. Do you see the device split narrow for work queries, or is it about the same as consumer search?

[VisibleAI @usevisibleai]

@axios Referral counts miss part of it: a lot of the choosing now happens inside the answer. In our scans one project management tool was named in all 15 buyer answers, and none of that shows up as a visit unless someone clicks. Should publishers track mentions too, not just referrals?

[VisibleAI @usevisibleai]

@Adweek @OpenAI If ads can't be targeted on what people tell ChatGPT, the unpaid answer stays where most of the brand shortlisting happens. Do you expect advertisers to move budget toward earning those organic mentions instead?

[VisibleAI @usevisibleai]

@Digiday Counting appearances is a good first step, but it can flatter a brand. In our scans one CRM was named in all 15 answers yet recommended in only 60% of them. Will the dashboard show who got picked when you didn't?

[VisibleAI @usevisibleai]

@seroundtable With ads above the AI answer, the organic shortlist underneath is what's left to earn. In one category we scanned, an agency was named in 11 of 12 answers yet recommended in only 9% of those mentions. Do the Bing ads repeat brands the answer already names?

[VisibleAI @usevisibleai]

@dickiebush Writing every word yourself is probably what keeps the AI slicing from turning into slop. One more place worth checking: a lot of readers now ask ChatGPT before they search. Do you ever look at what it says when someone asks about your topic?

[VisibleAI @usevisibleai]

Scans were run with VisibleAI's free scan (ChatGPT, 5 questions x 3 runs). Try it on your brand: https://t.co/zD7O1A3Cy4

[VisibleAI @usevisibleai]

Small SEO agency for SaaS: 90/100, named in 11 of 12 ChatGPT buyer answers.
Two of the best-known SaaS SEO agencies: 35 and 29.

But ChatGPT recommended the 90 in only 9% of those mentions. Named is not chosen.

Which agency would you expect it to pick first?

[VisibleAI @usevisibleai]

@buffer Your own data beats best-practice threads in AI search too: when we re-ran the same buyer questions in ChatGPT three times, one CRM's score landed anywhere from 43 to 73. Do you suggest a minimum sample before people trust what their AI coach tells them?