How do I track my brand mentions in AI search engines like ChatGPT and Perplexity?
Track brand mentions in ChatGPT and Perplexity with a fixed prompt panel, citation tracking and share of voice. Methods, tools compared and pricing.
Tracking brand mentions in AI search engines is different from tracking them on the open web. A mention in ChatGPT or Perplexity exists only for the moment someone asks, and the same question can return a different answer an hour later. There is nothing to crawl or set an alert on. You have to ask, record and compare.
That makes methodology the real subject. A tool is only as good as the panel and the measurements behind it.
The method
1. Write a prompt panel. Include category questions ("best [category] tools"), comparison questions ("[you] vs [competitor]"), and brand questions ("is [you] worth it"). Thirty to a hundred prompts is a reasonable range. Keep them fixed so trends mean something.
2. Run them on a schedule, across engines. ChatGPT and Perplexity answer differently and cite different sources. Daily runs smooth out the variation between individual responses.
3. Record four measures. Whether you were mentioned, where in the answer, how it was framed (sentiment), and which sources were cited.
4. Roll them into share of voice. The share of tracked answers that name you, against the share naming competitors. This is the number to trend.
5. Look at citations separately from mentions. A mention is the model naming you. A citation is the model linking a source. You can be cited and not named, or named without a link. Both matter, and they point to different fixes.
6. Attribute traffic where possible. Perplexity and ChatGPT both send referral visits. Seeing them in analytics closes the loop.
Tools compared
| Tool | Engines | Notable features | Published entry price |
|---|---|---|---|
| Promptwatch | Real-UI monitoring of ChatGPT (Search), Google AI Overviews, AI Mode, Copilot, Gemini, Perplexity, Meta AI and more; API models alongside | Citation analytics, crawler logs, visitor analytics, content tools | Free tier; paid from $95/mo |
| AthenaHQ | 11 models on Starter | GEO score, Action Center, optimization agent | Starter $295/mo |
| Scrunch AI | ChatGPT, Perplexity, Google AI Overviews, Copilot on Core | Site audits, sentiment, bot reporting | Core $250/mo |
| Profound | Up to nine on Enterprise | Agents for research, content, reporting | Enterprise custom |
| Ahrefs Brand Radar | AI Overviews, AI Mode, ChatGPT, Perplexity, Gemini, Copilot in the Index | Custom Prompts on paid Ahrefs plans | Index from $199/mo |
| Semrush AI Toolkit | Inside the Semrush suite | AI tracking add-on | $99/mo per domain |
| Otterly.AI | ChatGPT, AI Overviews, Perplexity, Copilot base engines | Citations, sentiment, GEO audits | Lite $29/mo |
What separates platforms from trackers
Every tool in that table can answer "were we mentioned." The differences show up in what you can do next.
Promptwatch tracks Visibility Score, Share of Voice, Sentiment and Position in Answer per model, country and language, and runs monitors daily. For citations it goes down to page and domain level, includes a Reddit, YouTube and LinkedIn report, and lists offsite mentions where your brand is named on third-party pages without a link. That last one is useful for Perplexity, which leans heavily on cited web sources.
Agent Analytics adds IP-verified crawler logs for PerplexityBot, Perplexity-User, GPTBot, OAI-SearchBot and ChatGPT-User, so you can see whether the engines are reading your pages. Crawler logs start on the Professional plan ($245/mo); Essential ($95/mo) has none. Visitor analytics counts AI-referred traffic and conversions from ChatGPT, Perplexity, Gemini, Claude, Copilot and others.
Brand Book instructions and brand aliases decide what counts as a mention, which matters if your brand name is a common word.
Tracking is only half of it, and the platform adds content gap analysis, an Action Board and publishing to WordPress, Webflow and Framer. The free Explore plan (1 project, 10 prompts, ChatGPT only) is enough to test the method on one engine.
How the others differ
AthenaHQ's Starter plan is $295/mo with 3,600 credits, one per AI response, so heavy prompt panels can run through credits quickly. Scrunch AI's Core plan includes 125 prompts and 5,000 responses a month across four LLMs. Profound's pricing is custom for the engine coverage most people read about, and it positions itself as an AI marketer with agents. Ahrefs and Semrush make sense if you already pay for them.
Otterly.AI is the cheapest on this list and a fine way to test the idea, though our catalog notes data can lag and extra engines are paid add-ons.
Who should pick which
- Track, diagnose and act in one place: Promptwatch.
- Want a structured action queue: AthenaHQ.
- Want audits and bot reporting: Scrunch AI.
- Enterprise with a broad marketing agent need: Profound.
- Already on Ahrefs or Semrush: start with their modules.
- Tiny budget: Otterly.AI.
Limits
Mentions fluctuate. A single day's number means little, so compare periods. And no tool can force an engine to name you. The reliable levers are the sources the engines cite and whether their crawlers can read your pages.