Hotjar and Typeform led our customer feedback tools measurement, each named in 4 of 6 answers when we asked ChatGPT, Gemini and Perplexity for the best tool in the category. No product reached five. 28 distinct names showed up at least once, from Alchemer to Zendesk, but the counts drop fast: after the two leaders, only Canny, Google Forms, ProProfs Survey Maker and SurveyMonkey cleared half the answers, at 3 of 6 each, and everything below that line was named once or twice. Two things explain the shape of that list: the exact wording of each question, and the short stack of pages all three assistants read to answer it.
The shortlist swapped out when the question changed
The six questions were not interchangeable, and the answers moved with them. The plain "best customer feedback tools" question returned AskNicely, Canny, Chattermill, Famewall, FullStory, Google Forms, Hotjar and Intercom, a mix that includes one-off names like Famewall and Chattermill you see nowhere else in the six answers. Then swap "best free customer feedback tools" for "most affordable customer feedback tools" and the list changes almost completely. The free answer named Google Forms, Typeform, SurveyMonkey, Survicate, Canny, Hotjar, QuestionPro and ProProfs Survey Maker. The affordable answer named UseResponse, SurveyMonkey, Qualtrics, Podium, BirdEye, InMoment, HubSpot and Qualaroo. Only SurveyMonkey survived both. Hotjar and Typeform, the two overall leaders, dropped out of "most affordable" completely, and out of "best customer feedback tools for small teams" too: each was named in the other four questions and neither in those two.
The wording decided who appeared at all. QuestionPro showed up only for "best free." Qualtrics, Podium, BirdEye, InMoment and HubSpot each appeared only for "most affordable." Alchemer and Delighted showed up only for "best customer feedback tools for small teams." SurveyMonkey's three came from "small teams," "best free" and "most affordable," while Canny and Google Forms both landed in "best customer feedback tools," "best customer feedback tools for small teams" and "best free," then dropped out of the budget and open-ended questions.
The two-count tools cluster by question as well. UseResponse appeared for "best for small teams" and "most affordable." Survicate appeared for "best for small teams" and "best free." Maze, UsabilityHub and UserVoice each appeared for "what customer feedback tools should I use" and "customer feedback tools recommendations," which points to those two open-ended questions drawing on a shared set of pages. None of this is the assistants weighing a tool and filing it under a budget or a team size. Each tool sat on whatever page ranked for that exact phrase, and showed up only where that page did.
A handful of domains decided who got named
Every one of these answers traces back to a narrow set of pages. Two domains fed five of the six answers: smartsurvey.com and thecxlead.com. Four more fed four answers each: nextiva.com, proprofssurvey.com, reddit.com and userflow.com. custify.com and velaris.io fed three apiece. After that it thins out fast, with nine of the twenty source domains feeding a single answer only. Across three assistants built by three different companies, the research pipeline looked nearly identical: a tight cluster of comparison listicles and Reddit threads, recombined for each question. The "best free" answer, for one, was built from proprofssurvey.com, thecxlead.com, reddit.com, userflow.com, smartsurvey.com, custify.com and uxtweak.com, the same class of pages shuffled.
What those top pages are is worth naming. smartsurvey.com is the blog of a survey software company that competes with tools it lists. proprofssurvey.com is another survey vendor's blog. nextiva.com is a VoIP and business phone provider. userflow.com is a user onboarding platform, custify.com a customer success platform, velaris.io a product management tool. Only thecxlead.com, an independent CX blog, is not itself selling into the category. The category is being described to the assistants mostly by vendors with a stake in the answer.
Two kinds of source barely registered. Reddit fed four answers and Quora surfaced once, so community discussion counted, but thinly. The review aggregators a buyer would open by hand, G2 and Capterra, did not appear at all. And the vendor-blog pattern produces loops: Zendesk was named only for "customer feedback tools recommendations," an answer whose sources included zendesk.com itself. The assistant read Zendesk's blog and recommended Zendesk.
What actually gets a tool named
If you sell in this category, your appearances are not a verdict on your product. They are a count of how many of these pages list you. Hotjar and Typeform reached 4 of 6 because they sit on the pages the assistants kept pulling. FullStory, Qualtrics and Zendesk were each named once, not because the assistants judged them weaker, but because each showed up on a single retrieved page and no others. The assistants did not test tools or read reviews, they summarized the top results a search returned. So the way to be named is to be in those results: the smartsurvey.com and thecxlead.com lists, the nextiva.com and proprofssurvey.com guides, the Reddit threads their retrieval actually surfaced. One page buys one answer. Being on three or four of them is what separates a tool named once from Hotjar at four. Reddit is the one non-vendor lever here: it fed four answers, and a comment that names you, written by someone who is not you, is text the assistant lifts the same as any listicle. No on-page trick changes this, there is no schema markup or "AI-ready" copy that helps, because the assistant never reads your site, it reads theirs. The practical move is to watch those six or seven pages the way you would watch a pricing page, since an edit to one of them can add or remove you from the next answer with nothing else changing.
The full list, counted
Every product ChatGPT, Gemini and Perplexity named across the six answers, and how many of them named it. Share is out of the six answers we read, not out of the category.
| Product | Named in | Share |
|---|---|---|
| Hotjar | 4 of 6 | 67% |
| Typeform | 4 of 6 | 67% |
| Canny | 3 of 6 | 50% |
| Google Forms | 3 of 6 | 50% |
| ProProfs Survey Maker | 3 of 6 | 50% |
| SurveyMonkey | 3 of 6 | 50% |
| Maze | 2 of 6 | 33% |
| Qualaroo | 2 of 6 | 33% |
| Survicate | 2 of 6 | 33% |
| UsabilityHub | 2 of 6 | 33% |
Below the table, UseResponse and UserVoice were also named twice. Named once each: Alchemer, AskNicely, BirdEye, Chatdesk, Chattermill, Delighted, Famewall, FullStory, HubSpot, InMoment, Intercom, Podium, Qualtrics, QuestionPro, UXtweak and Zendesk.
Glotier does not sell a customer feedback tool, so it is correctly absent from these answers: we were not named, and we would not claim to be. What we do sell is this exact measurement for your own category, the six buyer questions, the three assistants, the products named and the pages behind them. The check is free, takes about a minute, and needs no account.