Google Search Console was named in 6 of the 6 answers we measured, the one technical SEO tool no phrasing could dislodge. On 2026-08-04 we put six buyer questions about technical SEO tools to ChatGPT and Gemini, gave each one live web search, and recorded every product the answers named and every source domain behind them.
Two paid suites tied for second: Ahrefs and Semrush were each named in 4 of 6 answers (67%). Behind them came the dedicated crawlers Screaming Frog and Sitebulb, plus Google's free PageSpeed Insights, each named in 3 of 6 (50%). In total 22 different products appeared across the six answers, but only 12 of them were named more than once.
Room was tight in every answer. The six answers named between 6 and 8 products each (the "small teams" question named the fewest, 6), so landing on any list at all meant beating out a field where the assistants only had space for a handful.
How the shortlist changed across the six questions
The most useful thing we saw is that the roster almost completely swapped when the question moved from "best" to "free" or "affordable." The same category, worded three different ways, produced three different lists.
The three neutral phrasings behaved almost identically. "Best technical seo tools," "what technical seo tools should I use," and "technical seo tools recommendations" returned the same stable spine every time: Google Search Console, Ahrefs, Semrush, Screaming Frog, Sitebulb, and PageSpeed Insights. SE Ranking is the odd one out here, named in exactly 1 of 6 answers, the bare "best" query, and nowhere else.
The "best free" question threw out every paid suite. Ahrefs, Semrush, Screaming Frog, Sitebulb, and Lumar all disappeared from that answer. In their place came Google's free stack (Search Console, Trends, Keyword Planner, Autocomplete) and free tiers: Ahrefs Webmaster Tools, Moz, and SEO Review Tools. Four of those, SEO Review Tools, Google Autocomplete, Ahrefs Webmaster Tools, and Moz, were named in that single "free" answer and nowhere else in the six.
That answer is also where the double-labelling shows. Ahrefs the paid product was named in 4 of 6 answers, while Ahrefs Webmaster Tools, its free tier, was named in 1 of 6. We keep those as two separate labels rather than adding them into one number the measurement never produced.
The "most affordable" question swapped the roster again. It dropped Ahrefs and Semrush completely and pulled in budget names: Keysearch, SEOmator, SEO Site Checkup, Google Analytics, and Moz Pro Starter, each named in that one answer and nowhere else. Ubersuggest and Mangools bridged it to the "small teams" answer, each named in 2 of 6. Moz Pro Starter (1 of 6) is the same brand as Moz (1 of 6) under a second label, and we count them separately too.
Only Google Search Console survived all six phrasings. Everything else was a function of how the buyer worded the question.
The source map: who fed the answers
Behind those 22 products sat 23 source domains, and they were not review directories. One roundup blog, marketermilk.com, was cited in all 6 of the 6 answers. Reddit was second, feeding 5 of 6 (it missed only the "small teams" answer). Then morningscore.io and technicalseo.com at 4 of 6 each, and nimblechapps.com and zapier.com at 3 of 6. Community threads and roundup articles did nearly all of the work.
The classic software directories barely registered. Capterra did not appear once. G2 appeared a single time, and only through its blog arm learn.g2.com, not its product listings. Gartner appeared once. If you assumed AI shortlists in this category are built from directory star ratings, this data says the opposite.
The insight worth acting on: in four cases a vendor's own domain fed the exact answer that named its product. On the "best free" answer, moz.com was a source and Moz was named, ahrefs.com was a source and Ahrefs Webmaster Tools was named, and seoreviewtools.com was a source and SEO Review Tools was named. On the "most affordable" answer, seomator.com was a source and SEOmator was named. When the field opened up past the famous suites, the assistants reached for a vendor's own page and repeated its product back.
What a vendor in this category would do to get named
If we sold a technical SEO tool, this page would be our to-do list, and it would have little to do with market share. SEOmator is not a household name, yet it was named in the "affordable" answer because seomator.com and marketermilk.com were both in front of the assistant for that question. The source list decides who gets named.
So the work is concrete: be in the pages these two assistants actually read. That means the roundups that recur (marketermilk.com in 6 of 6, morningscore.io and technicalseo.com in 4 of 6, nimblechapps.com and zapier.com in 3 of 6) and the Reddit threads that fed 5 of 6 answers. Chasing a Capterra badge would have earned nothing here, because Capterra was cited zero times.
Your own domain can also carry you on the long-tail cuts. The "free" and "affordable" answers named 8 or fewer products and reached for vendor pages to fill out the list. It is not automatic, though: morningscore.io fed 4 of the 6 answers and Morningscore the product was named in 0 of them, because those pages were roundups of other tools. A page on your own site earns you a mention when it credibly names your product for the exact question a buyer asked.
The full list, counted
| Product | Named in | Share |
|---|---|---|
| Google Search Console | 6 of 6 | 100% |
| Ahrefs | 4 of 6 | 67% |
| Semrush | 4 of 6 | 67% |
| PageSpeed Insights | 3 of 6 | 50% |
| Screaming Frog | 3 of 6 | 50% |
| Sitebulb | 3 of 6 | 50% |
| Google Keyword Planner | 2 of 6 | 33% |
| Google Trends | 2 of 6 | 33% |
| Lumar | 2 of 6 | 33% |
| Mangools | 2 of 6 | 33% |
| Surfer SEO | 2 of 6 | 33% |
| Ubersuggest | 2 of 6 | 33% |
These are the 12 products named in two or more of the six answers. Ten more were named once each: SE Ranking, Ahrefs Webmaster Tools, Moz, Moz Pro Starter, Keysearch, SEOmator, SEO Review Tools, SEO Site Checkup, Google Analytics, and Google Autocomplete.
Glotier does not sell a technical SEO tool, so we are correctly absent from all six of these answers, and we would be worried if we were not. We are not going to claim an AI recommended us in a category we are not in. But running exactly this measurement, six buyer questions, the real products each answer named, and the real source pages behind them, is what Glotier does for whatever category you do sell in. The check is free, needs no account, and takes about a minute.