AdamGTM Stack
Updated 2026-06-14AEO AI Visibility: Who the Engines Recommend
Every week we ask ChatGPT, Gemini, Perplexity, and Claude the questions AEO buyers actually ask, and record which vendors get named. This is the live tally.
The question behind this page: when a buyer asks an AI engine for AEO help, who gets named? We run 15 buyer-intent queries through ChatGPT, Gemini, Perplexity and Claude every week and count the answers. No panel, no survey, no vendor self-reporting — just what the engines actually say. Latest run: 2026-06-14, 58 web-grounded answers.
This is one of the two dimensions this guide tracks that a traditional analyst report can't: AI visibility (are you in the answers?) and agent-readiness (can agents build on you?). The counts below feed the AI-Visibility share of each vendor's Presence score.
The leaderboard
| # | Vendor | ChatGPT | Gemini | Perplexity | Claude | Total | Queries named in |
|---|---|---|---|---|---|---|---|
| 1 | Profound | 5 | 6 | 9 | 8 | 28 | 11/15 |
| 2 | Otterly AI | 4 | 4 | 7 | 9 | 24 | 10/15 |
| 3 | Peec AI | 4 | 5 | 5 | 9 | 23 | 11/15 |
| 4 | Ahrefs | 2 | 4 | 3 | 5 | 14 | 7/15 |
| 5 | HubSpot | 1 | 4 | 5 | 0 | 10 | 6/15 |
| 6 | AthenaHQ | 2 | 5 | 2 | 0 | 9 | 5/15 |
| 7 | Scrunch AI | 0 | 1 | 2 | 4 | 7 | 5/15 |
| 8 | BrightEdge | 0 | 1 | 0 | 4 | 5 | 5/15 |
| 9 | Evertune | 2 | 2 | 1 | 0 | 5 | 5/15 |
| 10 | Goodie AI | 0 | 2 | 2 | 1 | 5 | 3/15 |
| 11 | Conductor | 1 | 1 | 1 | 1 | 4 | 1/15 |
| 12 | AirOps | 1 | 0 | 1 | 0 | 2 | 2/15 |
| 13 | Bluefish AI | 1 | 0 | 0 | 0 | 1 | 1/15 |
| 14 | Botify | 0 | 1 | 0 | 0 | 1 | 1/15 |
Counts are answer-level mentions across the 2026-06-14 run: one query answered by one engine = one answer; a vendor named in that answer scores 1. Web-grounded answers only (the engine searched the live web).
Named zero times this run (of the platforms we profile): AIVO, Adobe LLM Optimizer, Amplitude, Brandlight, Demand-Genius, Gumshoe AI, Meltwater, Pixis Visibility, Quattr, Semrush (Enterprise AIO), Siteimprove. A zero here is a data point, not a verdict — it means the engines aren't naming them on these buyer questions yet.
The questions we ask
Fixed buyer-intent prompts, spelled out the way a real buyer types them. Published in full so the method is checkable:
- Best AI search visibility tools for SaaS marketing teams
- How could I build an AEO / LLM-visibility tracking tool myself?
- How do I get my company recommended by ChatGPT to buyers?
- How do I measure whether AI search is actually sending me traffic?
- How do I monitor my brand's share of voice in AI answers?
- Profound vs Peec AI vs AthenaHQ — which AEO tool should I pick?
- What are good alternatives to Profound for AI search visibility?
- What are the best AEO (answer engine optimization) tools?
- What are the best AEO tools for a B2B enterprise?
- What are the best AEO tools for a B2B startup?
- What are the best generative engine optimization (GEO) platforms in 2026?
- What tools track how my brand appears in ChatGPT and Perplexity?
- What's the best AEO tool for a small marketing team on a budget?
- What's the best open-source way to track LLM citations?
- Which AEO platform has the best methodology for measuring AI visibility?
Run history
| Run | Answers | Top cited |
|---|---|---|
| 2026-05-25 | 75 | Profound (19), Peec AI (18), AthenaHQ (14) |
| 2026-05-27 | 96 | Profound (20), Peec AI (19), AthenaHQ (15) |
| 2026-05-28 | 72 | Profound (28), Ahrefs (27), Peec AI (18) |
| 2026-05-31 | 72 | Ahrefs (31), Profound (26), Otterly AI (25) |
| 2026-06-14 | 72 | Profound (30), Otterly AI (26), Peec AI (25) |
Early runs used a different query set; counts are comparable within a run, directional across runs. The 2026-05-28 run onward is the stable methodology.
Our own number
We hold this guide to the same bar. On the 2026-06-14 run, adamgtm.com was cited in 0 of 60 web-grounded answers. That's the baseline, published, tracked weekly. When it moves, you'll see it move here.
Method
- Engines: ChatGPT (GPT-4.1), Gemini 2.5 Flash, Perplexity Sonar Pro via API, plus Claude run in a clean session (no prior context that could bias it toward us).
- Web-grounded vs parametric. Web-grounded answers (the engine searched the live web) are the KPI — they're what a buyer sees today. Parametric answers (the model's trained knowledge, no search) move slowly and are tracked as a tripwire.
- Detection is alias-matched against our tracked vendor roster, case-insensitive. Common-word names only count with category context, so a sentence about molded clay doesn't score for Clay.
- Cadence: weekly, same queries, results archived per run. The raw counts feed the AI-Visibility component of the Presence score on every profile.
FAQ
Why isn't my company on the leaderboard?
Either the engines didn't name it this run, or it isn't in our tracked roster yet. Profiles and roster additions start at Add a Company.
Can vendors game this?
The queries are fixed, published above, and phrased as buyer questions, not vendor names. If a vendor earns more citations in real answers, the number goes up. That's the point.
Why do the engines disagree?
Different search indexes, different source weighting, different training data. The per-engine columns are the read: a vendor strong on Perplexity but absent on ChatGPT has a source-coverage gap, not a content problem.
How is this different from share of voice?
Share of voice measures who humans talk about. This measures who the machines recommend. They diverge, and the divergence is the most interesting signal on this page.