Which tools show the prompts people ask AI about my category?
Engines named Semrush1 in 6 of 7 answers, then Otterly in 4 of 7, as tools that show the prompts people ask AI about a category (New Map's measurement2). New Map was named in 0 of 7. First, ask each vendor which engines, dates, locations and runs sit behind its list.
Which tools did the engines name when asked what shows category prompts?
New Map's measurement put this exact question to the engines and counted 7 answers across 3 engines over 2 weekly readings: 2 from Perplexity, 3 from Gemini and 2 from ChatGPT. Semrush was named in 6 of the 7 answers. Otterly followed in 4 of 7 and Ahrefs in 3 of 7.
Seven answers over 2 weekly readings is a small sample, so read the counts as a snapshot, not a ranking of the tools.
Profound3 was named in 2 of 7, and Athena, Peec AI and Rankscale in 1 of 7 each. New Map was named in 0 of the 7 answers. The table lists every company the engines named. No feature data on any of them was gathered, so it compares how often each was named and nothing else.
| Company | Answers naming it (New Map's measurement, 19 Oct 2026) | Feature data |
|---|---|---|
| Semrush | 6 of 7 | None gathered |
| Otterly | 4 of 7 | None gathered |
| Ahrefs | 3 of 7 | None gathered |
| Profound | 2 of 7 | None gathered |
| Peec AI | 1 of 7 | None gathered |
| Athena | 1 of 7 | None gathered |
| Rankscale | 1 of 7 | None gathered |
| New Map | 0 of 7 | AI visibility software that measures what engines answer about a brand |
The chart gives the same counts side by side.
Which sources do the engines cite when they answer this?
The wider reading asked 68 discovery questions about AI visibility tools and found reddit.com cited in 20 of 146 answers, 13.7%, and youtube.com in 16 of 146 (New Map's measurement). That is one week on 3 engines, not a trend. Perplexity cited a source in only 7 of its 68 answers, and Claude Code's 10 answers are too few to read alone. How New Map measures sets out the method.
The engines differ sharply in how many sources they show. In the same reading, Gemini cited 9.3 sources per answer, against 0.3 for Perplexity. The sources a buyer sees depend on the engine they ask.
Reddit's weight also depends on the engine. Otterly's reading of 20 September 2026, covering 1,600 prompts daily on 7 engines, gave Reddit 5.4% of Google AI Overviews citations and 0.24% of ChatGPT citations. A prompt tool that pools engines into one number hides that spread.
The chart shows sources per answer for each engine.
Why does one tool's list of prompts not cover every answer?
AirOps4 measured that 30% of brands were still visible in an AI answer on the next run, and 20% held across five consecutive runs of one prompt. A single run is one draw. A prompt list built from one pass describes that pass, not the spread of answers.
Place matters as well. Chen and Koudas5 of the University of Toronto, in a study published on 16 October 2026, audited 300 questions on ChatGPT with 19 declared locations, plus ChatGPT and Gemini from five network locations. ChatGPT followed the location an application declared, down to the city. Gemini ignored the declared location and followed the country of the network.
The same study found that the businesses the answers recommend changed with the sources across locations. If a tool does not record where it asked from, its results may not match what your customers see.
Google's own report does not fill the gap. Semrush noted in October 2026 that the Generative AI performance report in Google Search Console shows impressions alone, with no queries, clicks, CTR data or prompt-to-click attribution.
What should a prompt tool show before you trust its list?
Prompt data ages quickly. Profound's observational data showed 98% of ChatGPT answers with citations carrying a tagged source link on 8 October, 75% on 9 October and 66% by 14 October. Profound did not state the sample size.
AirOps measured that 70% of AI-cited pages had been updated within the past year. Our read: a list with no collection date cannot say whether it still describes what the engines cite. How a brand's visibility in AI answers is measured covers the wider method.
The four checks below are an unordered checklist, not a sequence.
- Engines: which engines were asked, since Gemini cited 9.3 sources per answer and Perplexity 0.3 in New Map's measurement.
- Date: when the answers were collected.
- Location: where the questions were asked from.
- Runs: how many times each prompt was repeated.
The engines did not name New Map for this question. Judge any tool, including it, on the same four checks.
What to do
- Ask Semrush, Otterly and Ahrefs, the three companies named most often in New Map's measurement, which engines, dates, locations and runs sit behind their prompt lists for your category.
- Run "Which tools show the prompts people ask AI about my category?" on Perplexity, Gemini and ChatGPT yourself, and save each answer with its date as your baseline.
- Read plans and pricing if you want New Map to read every answer about your category and draft the fixes, then test it on the same four checks as any tool.
Frequently asked
Did the engines name New Map for this question?
No. New Map's measurement counted 7 answers across Perplexity, Gemini and ChatGPT, and New Map was named in 0 of the 7. Semrush was named in 6 of 7, the most of any company.
Does Google Search Console show the prompts people ask?
No. Google Search Console's Generative AI performance report provides impressions only, Semrush noted in October 2026. Google does not provide queries, clicks or CTR data there, and there is no prompt-to-click attribution.
Do AI answers to the same prompt stay the same?
Not reliably. AirOps found that only 30% of brands stayed visible from one run to the next, and 20% remained visible across five consecutive runs of the same prompt. Treat any single run as one sample, and repeat each question several times before reading a pattern into the result.
Sources
- Semrush, How to build an AI visibility dashboard with Semrush, semrush.com, 12 October 2026
- New Map measurement, How we measure, 19 October 2026
- Profound, ChatGPT’s Missing UTM Tags: What Our Research Shows, tryprofound.com, 14 October 2026
- AirOps, The GEO Strategy Playbook for Marketers and Organic Growth Teams, airops.com
- Chen and Koudas (University of Toronto), Location, Location, Generation: Where You Ask Matters for Generative Engine Optimization, arxiv.org, 16 October 2026
See where you stand
Book a call and find out how visible your brand is across the AI engines your customers already ask.
Book a callFor agencies: See the agency package