1001 SEO Media
All posts
By 1001 SEO Mediaai-generated answerscontent trackingcitationsagency practice

How do I track whether my content appears in AI-generated answers?

How we track whether a client's articles appear in AI-generated answers: fetched, cited and clicked, page by page, with a refresh routine for decaying pages.

Clients usually ask whether their brand shows up in AI answers. The question we care more about is whether their content does. A brand can be mentioned without a single owned page being used, and a page can be cited without the brand name in the answer. Content tracking follows individual URLs, and it changes how we run a program.

Three events, tracked separately

We keep a simple record for every important page.

Fetched. An AI crawler requested it. Without this, nothing else happens. We look at the bots by name: GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-User, PerplexityBot, Perplexity-User and Google-Agent.

Cited. An answer listed the page as a source. Engines often fetch more than they cite, which is why the GEO Inspector Chrome extension is useful for spot checks: it shows ChatGPT fan-outs and fetched versus cited sources, and runs locally.

Clicked. A visitor arrived from the assistant. We see that in visitor analytics.

If a page is fetched and never cited, it is a content problem. If it is never fetched, it is an access problem. We say that sentence to clients often because it prevents a lot of wasted rewriting.

The tool we use for it

We run content tracking in Promptwatch. These are the pieces we use most.

  • Citation analytics gives page and domain level citations with URL positions, so we can see where a client's page sits among the sources.
  • Page Tracker tracks specific URLs, ours or a competitor's, and links them to the prompts and responses where they appear.
  • Citation Trends classifies every citation by content type and source type, and reports citation rate over time with first cited, ramp-up days, peak rate and decay half-life per page.
  • Agent Analytics shows IP-verified crawler logs, a crawl-to-citation path with a citation rate per page, and errors with failure explanations. Filters are stackable and results export to CSV.
  • 360 Insights joins crawler logs, citation clicks and AI-referred visits per page, with Topic Insights showing where a page underdelivers for its topic.
  • Visitor analytics attributes AI-referred traffic and conversions.

Our weekly routine

  1. Pull the page list. Our top 20 to 50 pages per client, each with the question it should answer.
  2. Check crawl status. Any key page with no recent AI crawler hits or with errors goes to the top of the list.
  3. Check citations. Which pages gained, which lost, and which are new.
  4. Read the decay curve. Citation Trends shows how long each page took to ramp up and how fast it decays. We use that to decide when to refresh.
  5. Compare with rivals. Page Tracker on a competitor page that keeps beating ours tells us what to match.
  6. Act. Content Refresh gives diff-style update suggestions for existing pages, and the Action Board holds the prioritized list.

What we have learned to look for

New pages take time. The ramp-up days metric stops us from judging a new article too early.

Old winners fade. Decay half-life is the number that turns "this used to be cited" from a hunch into a task.

Format matters. The content type breakdown shows whether comparison pages, guides or list posts win in a client's category, so we write more of what the engines actually cite.

Off-site is part of it. Some of the citations that matter sit on other sites. Offsite mentions, and the Reddit, YouTube and LinkedIn reports, show where the client is named without a link.

When we publish new content

For clients on Business and Enterprise, the Content Agent plans, writes and publishes against visibility gaps on a schedule, with a review inbox or a hands-off mode with a quality check. We use the review inbox. It publishes to WordPress, Webflow and Framer. Framer has no per-item publish, because going live redeploys the whole Framer site. For every client, on-demand content creation from a brief and content gap analysis with coverage scores are available.

Plan notes

The free Explore plan covers 10 prompts on ChatGPT only. Essential at $95 a month adds all models, 50 prompts and 500 Agent Credits, but no crawler logs. For page-level work, we recommend Professional at $245 a month, where crawler logs begin and which includes 25M crawler logs and a 100-item knowledge base. Agency plans (Kick-off $199, Growth $399, Scale $799) include crawler logs and unlimited projects.

What we tell clients

We cannot make a page get cited. We can tell you whether it is fetched, cited and clicked, show how long it takes to ramp up and fade, and point to the fix. Start with the 20 pages that matter and a two-week baseline before you change them.