Are We Visible in ChatGPT? What Agencies Can Measure First
The board slide says "AI visibility" in bold. The client wants a yes or no: are we showing up in ChatGPT, Perplexity, or Google AI Overviews? Your team wants an answer that does not require buying another SaaS category before Friday.
In delivery we split the ask into layers. Citation tracking (prompt libraries, GEO dashboards) is probabilistic: useful for trends, weak as courtroom proof when the prompt set is generic. Deterministic work is different: robots.txt for GPTBot and related bots, HTTP status on priority URLs, lab LCP/INP/CLS on the pages sales and docs actually cite, and alerts when a deploy breaks fetchability.
Mixing those layers on one slide creates week-two arguments when checkout is slow but the homepage prompt chart looks fine. Name the layer first, then pick metrics that match it.
Confirm what "visible" means in the SOW: citations, crawl access, traffic impact, or all three.
Audit production robots.txt for GPTBot and related AI user-agents; document opt-outs in writing.
Baseline ten to twenty priority URLs (pricing, docs, key templates), not the whole sitemap.
Schedule lab tests and budgets on that list before you procure GEO tooling for citation charts.
Apogee Watcher covers the deterministic side across client portfolios. We do not score ChatGPT mentions. We tell you when priority URLs slow down, error, or drift after a deploy that might affect crawl access. Layer GEO when the retainer includes prompt-level citation reporting.
Read more: AI search visibility for agencies: what to measure first
