Heard of You visibility log

Our own AI visibility, measured monthly with 27 fixed prompts on 5 engines and published unedited. Baseline October 2026: not yet run; we expect near zero.

By Dasha. Updated November 1 2026. https://heardofyou.io/research/heard-of-you-visibility-log

This is Heard of You's own AI visibility. We measure it the way we measure clients', and publish it every month, flattering or not.

We ask ChatGPT, Claude, Gemini, Perplexity and Google's AI Overviews and AI Mode 27 fixed questions. They are questions a founder might ask when looking for a service like ours. Each runs three times, and we record how often we are named.

The October 2026 baseline has not run yet. We expect it at or near zero. The site went live on October 31 2026, and no third-party page mentions us yet. The table below has one row so far.

Method
- Prompt set: 27 prompts, fixed. Prompts 1 to 25 are "Heard of You's own competitor prompts" from the research we launched the site on, unchanged. We added 26 and 27 on October 31 2026, for the two closest competitors missing from the original set. A prompt is never edited. New ones get new numbers.
- Engines (5): ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews / AI Mode. We weight Claude and ChatGPT when reading results. Founders pay for Claude more than any other assistant (59%, Supabase) and use ChatGPT most (72.1%, Lenny's Newsletter).
- Runs: 3 per prompt per engine, in a fresh, signed-out chat, with a United States location set. Location changes answers (Chen and Koudas, 2026). We record the model version shown. 27 x 3 x 5 = 405 answers per full run.
- Why three runs: the same prompt returns the same brand list less than 1 time in 100 (SparkToro). One answer is not a measurement.

Cadence

| Run | What | Answers | When |
|---|---|---|---|
| Baseline (day 0) | All 27 prompts, 3 runs, 4 assistants by API, plus Google AI Overviews by hand | 324 by API + 81 by hand | Friday November 6 2026, before any outreach; published Monday November 9 |
| Daily pulse | The 5 starred prompts (1, 4, 7, 10, 20), 1 run, 4 assistants | 20 a day | Every morning at 7:10 ET, internal; the weekly average is published as one line in the monthly update |
| Weekly full | All 27 prompts, 3 runs, 4 assistants, plus 10 branded prompts about Heard of You itself, 1 run | 364 | Every Monday, internal |
| Monthly published | The first weekly full of the month plus the 81 Google AI Overviews answers by hand | 445 | First Monday of each month, published here |
| Day 30, 60, 90 checkpoints | The monthly published rows, compared in the same sheet we use for clients | | December 7 2026, January 4 2027, February 1 2027 |

What we record for every answer
- Named (yes or no)
- Recommended versus only cited
- Position words ("top pick", "also worth a look")
- Competitors named
- Sources cited, with URL
- Whether one of our own pages was cited, and which
- Tone
- For the branded prompts 23 to 25: whether what the assistant said about us (prices, what we do) was accurate

The headline number is the mention rate: the share of answers that name us, per engine and overall. We never report a rank.

Results

| Run | Date | Answers | Named (overall) | ChatGPT | Claude | Gemini | Perplexity | Google AI | Our page cited | Named instead (most often) | Notes |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Baseline, October 2026 | not yet run | 0 of 405 | not yet run | | | | | | | | Site launched October 31 2026. Baseline scheduled for the first week of November, before any outreach. We expect at or near zero. |

Each monthly row links to its raw run log (CSV, CC BY 4.0) once published.

The 27 prompts

Shapes follow how real people prompt: short, with "best", a budget or a personal detail (Stella Rising). ★ marks the five weekly-pulse prompts.

| # | Shape | Prompt |
|---|---|---|
| 1 ★ | Best for use case | best way to get my SaaS recommended by ChatGPT |
| 2 | Best for use case | AI visibility service for solo founders |
| 3 | Best for use case | who can help my indie app show up in AI answers |
| 4 ★ | Price-capped | AI visibility audit under $200 |
| 5 | Price-capped | cheap done-for-you GEO for a bootstrapped startup |
| 6 | Price-capped | is there a one-time AI visibility fix, no monthly retainer |
| 7 ★ | Persona | I'm a solo founder and ChatGPT never mentions my product, what should I do |
| 8 | Persona | indie hacker, $2k MRR, how do I get Claude to recommend my tool |
| 9 | Persona | I'm not a marketer, who can get my app into AI recommendations for me |
| 10 ★ | Alternatives | alternatives to Peec AI for a one-person company |
| 11 | Alternatives | Profound alternatives that are cheaper |
| 12 | Alternatives | Otterly alternative that actually does the work for you |
| 13 | Alternatives | something like FirstIndex but for indie makers |
| 14 | Alternatives | is there a cheaper alternative to AI Find My Business |
| 15 | Comparison | Heard of You vs Kovmere |
| 16 | Comparison | Peec vs Otterly vs Semrush AI toolkit for a small startup |
| 17 | Comparison | AI visibility tracker or done-for-you service, which is better for a solo founder |
| 18 | Comparison | ListingBott vs earned listicles for AI visibility |
| 19 | Comparison | best free AI visibility checker (pickedby.ai, HubSpot, Ahrefs) |
| 20 ★ | Problem first | why doesn't ChatGPT recommend my product |
| 21 | Problem first | how do I get mentioned in "best X" lists that AI cites |
| 22 | Problem first | does submitting to directories help AI visibility |
| 23 | Branded check | Heard of You reviews |
| 24 | Branded check | what is heardofyou.io |
| 25 | Branded check | is Heard of You legit |
| 26 | Alternatives | alternatives to Airefs |
| 27 | Alternatives | Siteline alternative |

The competitors in these prompts are real companies. They are there only because a founder comparing services would name them. We have no relationship with any of them.

What we expect, honestly

Near zero at baseline on prompts 1 to 22. No third-party page mentions us yet. Absent from the assistant's live search almost always means absent from the answer.

On the branded prompts 23 to 25, we expect the assistants to either not know us or to describe the homepage. What we check is whether what they say is accurate.

The interesting part is the monthly change. Does it track what we do: the research we publish, the places that write about us, the reviews from pilot clients? If it does not move, that goes on this page too.

What to do next

Run the same method on your own product. The do-it-yourself guide (/guides/check-if-chatgpt-recommends-you) walks through it. The prompt library (/prompts) has the templates. The free check (/check) does the first five prompts for you.

Questions

Why would you publish a zero?

A company that sells AI visibility and hides its own numbers is asking you to trust a pitch. A zero with a date on it is a starting line. If the number does not move, you will see that too.

Will you ever change the prompts?

We never edit one. New prompts get new numbers, so every month compares against the same questions. The two added so far, 26 and 27, are listed below.

Why five engines here but four in the free check?

The free check covers ChatGPT, Gemini, Claude, Perplexity and Muse. For ourselves we add Google AI Overviews and AI Mode. Founders looking for this kind of service still start in Google.


Sources
- SparkToro, AIs are highly inconsistent when recommending brands (https://sparktoro.com/blog/new-research-ais-are-highly-inconsistent-when-recommending-brands-or-products-marketers-should-take-care-when-tracking-ai-visibility/)
- Supabase, State of Startups 2026 (Claude is the top paid assistant among builders) (https://supabase.com/state-of-startups)
- Lenny's Newsletter, AI tools survey (ChatGPT most used among founders, 72.1%) (https://www.lennysnewsletter.com/p/ai-tools-are-overdelivering-results)
- Chen and Koudas, location effects in AI recommendations, 2026 (pre-print) (https://arxiv.org/abs/2610.24081)
- Stella Rising via Search Engine Land, how real people prompt AI (n=524) (https://searchengineland.com/real-people-actually-prompt-ai-geo-479731)