Traditional rank tracking doesn’t work anymore, at least not on its own. When a buyer asks ChatGPT “what’s the best CRM for a 20-person startup” and gets a generated paragraph with three tool names in it, there’s no ranking position to screenshot, just an answer that either includes you or doesn’t. Tracking brand mentions in AI search results means monitoring whether, how often, and in what context your brand shows up across ChatGPT, Perplexity, Gemini, and Google AI Overviews. Here are nine ways to actually do it, from free manual methods to paid platforms built for the job.
1. Run manual spot-checks with real buyer prompts
The free version of this costs about an hour a month: write 10 to 20 prompts a real buyer would actually type, the kind that shows commercial intent rather than generic curiosity, and run them through ChatGPT, Perplexity, and Gemini by hand. Log whether you’re mentioned, cited as a source, or actively recommended. It doesn’t scale, but it’s the fastest way to get a feel for where you stand before spending money on a tool.
Keep the log in a simple spreadsheet with one row per prompt per engine per month. It’s tedious at first, but after two or three months you’ll have enough history to see a trend instead of just a single noisy snapshot, which matters because AI-generated answers can vary between runs even for the same prompt.
2. Use a dedicated platform to track brand mentions in AI search
Once manual checks get tedious, a dedicated tool automates the same idea at scale: it runs hundreds of prompts against multiple AI engines on a recurring schedule and logs the results. Otterly.ai is the accessible entry point, with a Lite plan around $29 a month for 15 tracked prompts, scaling to $189 for 100 prompts with API access. Peec AI starts around $99 a month for basic competitive tracking and is a solid mid-market pick. Profound sits at the enterprise end with custom pricing, aimed at teams that need prompt-volume data to justify budget internally.
3. Separate mention, citation, and recommendation
Not all AI visibility is equal, and lumping it together hides what’s actually happening. A “mention” just means your brand name appears somewhere in the answer. A “citation” means the model linked to one of your pages as a source. A “recommendation” means the model actively suggested you as the answer to the buyer’s question. Track these three separately, because a page that gets cited constantly but never recommended is telling you something different than one that’s recommended but never cited.
4. Monitor which competitors show up in the same prompts
Your visibility numbers mean little in isolation. If you’re mentioned in 40% of relevant prompts but a competitor is mentioned in 85%, that gap is the actual story, not your raw percentage. Most dedicated tracking tools let you add competitor names to the same prompt set so you see both sides in one dashboard, which is far more useful than checking your own brand and a competitor’s separately.
5. Build a lightweight tracker on the Perplexity or OpenAI API
If you’re technical and want more control than a SaaS dashboard gives you, both Perplexity and OpenAI expose APIs you can script against: send a fixed set of prompts on a schedule, log the raw responses, and search them for brand and competitor names. It’s more setup work than a paid tool and you lose the dashboard, but it’s cheaper at scale and gives you full access to the raw text for further analysis.
6. Track sentiment, not just presence
Being mentioned isn’t automatically good. AI models sometimes surface a brand alongside a specific complaint, an outdated feature comparison, or a “but competitors X and Y are cheaper” qualifier that undercuts the mention. Whichever method you use, actually read a sample of the full responses rather than just counting whether your name appears; a rising mention count paired with consistently negative framing is worse than fewer, cleaner mentions.
7. Identify which sources the models cite most
When an AI engine cites a source in its answer, that source had outsized influence on what got said. Track which third-party sites and your own pages get pulled most often across your prompt set. If a competitor’s blog post keeps getting cited as the source for comparisons in your category, that’s a specific, fixable target: either get your own comparable content published and structured well enough to compete for that citation, or reach out about a correction if the cited content is inaccurate.
8. Set calendar reminders instead of relying on memory
AI search visibility shifts faster than traditional SEO rankings because model providers update retrieval behavior and underlying models more frequently than Google updates its core algorithm. A monthly check that felt sufficient at the start of the year can miss a real shift by the time you notice something feels off. Treat this like any other recurring metric: put it on a calendar, not on the list of things you’ll get to eventually.
9. Review and refresh your prompt set quarterly
The prompts buyers actually type change as your category evolves and as new competitors enter it. A prompt set built a year ago might be missing an entire class of questions buyers are asking now. Revisit your list of tracked prompts every quarter, retire ones that no longer reflect real buyer language, and add new ones based on actual sales conversations or support tickets, not just guesses about what people might ask.
One more thing worth naming: don’t expect these numbers to be perfectly stable week to week. AI models are probabilistic, and the same prompt run twice can surface different tools even with no changes on your end. Look at trends over four to eight weeks of data before drawing conclusions from any single check.
None of these nine methods is exclusive. A reasonable setup for most small teams pairs manual spot-checks early on with a mid-tier tool like Otterly or Peec AI once the category is worth tracking closely, and layers in competitor monitoring and sentiment review from day one rather than treating them as later add-ons. If you’re also working on the content side of AI visibility, our breakdown of AEO strategies that got us cited by Perplexity covers the tactics that move the mentions you’re now tracking.