Study · Building in public

We asked AI for the best AI visibility tool. Here is who it named, and who it did not.

We build a tool that measures whether AI names your product, so the honest thing is to run our own category through it and publish what comes back, flattering or not. We put the questions a buyer would actually ask, “best AI visibility tool”, “the cheapest one under $50”, “which do Reddit users recommend”, to ChatGPT, Gemini and Perplexity as live searches, and recorded who each named and the pages the answers were built from. The short version: there is no leader, the answers lean on a handful of sources, and AI named us zero times.

~12

tools AI named for our category

0

times it named us

95%

of answers Reddit fed, in our 40-question study

No leader. AI named a dozen different tools.

In a single run of our own category, the three assistants named a shifting set of about twelve products, among them Rankscale, Frase, Profound, Peec AI, PromptOpti, Temso, Nightwatch, GrackerAI, LLMClicks and LLM Pulse. What stood out was not the list but the disagreement: almost no tool was named by all three assistants. Most were named by exactly one. That matches the larger pattern from our 40-question study across other categories, where 223 of 266 named products appeared only once and the most-named product showed up in just three answers. No name owns “AI visibility tool” yet, which is unusual and, if you are building here, useful: a category with no leader is a category still up for grabs.

We wrote an honest, both-sides comparison of each of these against Glotier, where they win and where we do, with real pricing, for example Glotier vs Profound, vs Frase and vs Rankscale. The full set is in the roundup.

The answers came from a small set of sources, not the tools’ own sites.

When we recorded the pages each answer was built from, the same domains kept returning: SE Ranking’s AI-visibility page, Reddit threads, Frase, GrackerAI and Zapier. A short list of listicles and community threads, not the vendors’ marketing pages. In our broader 40-question study, Reddit was among the sources in 38 of 40 answers, and the review directories companies pay thousands a year for fed 0.7% of citations. The lesson repeats: assistants build answers from what they can retrieve and trust, and for tools that means a few “best X” roundups and the Reddit threads where real people compare options. Being genuinely the best product matters far less than being present in those specific pages.

AI named us zero times. We are publishing that.

In this run, Glotier was named in none of the questions. Not ranked low, absent. We could hide that, but the whole point of the tool is to show a company exactly this number, so hiding our own would be dishonest. What we are doing about it is the same thing we would tell any customer: get into the sources the assistants actually read. We wrote comparison and alternative pages for each tool AI named, we are working to earn mentions in the Reddit threads and roundups that fed those answers, and we re-measure weekly to see whether the number moves. Getting cited takes time, our own case study puts it at 66 days for the first five ChatGPT citations. This is week one of running that loop in our own category, in public.

What this means if you are trying to get named

Three things carry to any category. First, check whether there is a leader. If no single name dominates, the recurring-source set is small and getting into it is faster than fighting an entrenched incumbent. Second, find the exact pages the assistants pull from for your buyer questions, usually a few listicles and the Reddit threads where your category is discussed, and work to be present in those, not on your own blog. Third, measure it, because the sources are stable enough week to week that you can tell whether a change worked rather than guessing. That is the loop, and you can run the first step of it free.