How to track your citations in ChatGPT and Perplexity

Rankings don't tell you this

A page can rank well on Google and still never get cited in an AI-generated answer, because the two systems retrieve and select sources differently. If citation tracking isn't a separate, deliberate exercise, you won't know whether your GEO and AEO work is doing anything until it shows up in referral traffic, which is a slow and noisy signal.

Build the prompt list from real buyer questions

Generic prompts about your industry produce generic, unstable results. Prompts phrased the way an actual prospect would ask, including comparison and recommendation-style questions, are what actually matters for whether you show up when it counts.

• "What's the best [category] for [specific situation]?"
• "How does [your product] compare to [a named competitor]?"
• "What should I look for when choosing [category]?"

Run the same set on a schedule

Consistency matters more than frequency. Running the same 20 to 40 prompts monthly, across ChatGPT, Perplexity and Google AI Overviews, produces a trend line you can actually act on. Ad hoc, one-off checks don't tell you whether a change you made actually moved anything.

What to do with the gap list

When a competitor is cited and you're not, look at what their page does that yours doesn't: is the answer stated more directly, is the schema more complete, does it include a specific fact yours is missing. That comparison is a more reliable content roadmap than guessing at topics from a keyword tool alone.