Skip to main content

    Prompt Panel

    A prompt panel is a fixed set of 15–25 category-defining prompts — spanning definition, vendor-shortlist, comparison, and buyer-question intents — that is run monthly across ChatGPT, Perplexity, Gemini, Claude, and Grok to benchmark a brand's AI visibility over time. The panel stays constant so that month-over-month deltas are meaningful. Why it matters: Prompt panels are the LLM-era replacement for the keyword-rank tracker. They convert AI visibility from anecdote into a trend line, surface which engines and prompt classes need work, and prove — or disprove — that citation-building investments are compounding into share-of-model.

    Why Prompt Panel matters

    You can't manage what you don't measure, and AI visibility is invisible without a fixed prompt set. A Prompt Panel is the AEO equivalent of a keyword rank tracker — the baseline that lets you say "we moved from 12% to 34% share-of-model this quarter."

    In practice

    Build 15–25 prompts covering: brand queries, category queries ("best X for Y"), comparison queries ("X vs competitor"), and problem queries ("how do I solve Z"). Run monthly across ChatGPT, Perplexity, Gemini, Claude, and Grok. Log citations, rank order, and sentiment. Chart share-of-model over time.

    Common mistake

    Changing the panel every month. The whole point is comparability. Lock the panel for at least 6 months; add new prompts but never remove old ones during that window.

    How it connects

    The Prompt Panel is how the AEO Score gets ground truth and how LLM SEO progress becomes measurable. It's the input layer for share-of-model reporting.

    Frequently Asked Questions

    What is Prompt Panel?

    In short: Prompt Panel is a prompt panel is a fixed set of 15–25 category-defining prompts — spanning definition, vendor-shortlist, comparison, and buyer-question intents — that is run monthly across ChatGPT, Perplexity, Gemini, Claude, and Grok to benchmark a brand's AI visibility over time. See the full definition above for context.

    How many prompts should a Prompt Panel contain?

    15–25 is the sweet spot. Under 15 and results are too noisy (single-model variance dominates). Over 25 and monthly runs become expensive and slow. Weight the mix toward category and comparison prompts — those drive pipeline.

    Should I run the same prompts on every engine?

    Yes for the core panel — comparability across engines is the whole point. Add engine-specific prompts (e.g., 'cite three sources for X' works better on Perplexity than ChatGPT) as a secondary layer, not the baseline.

    How do I calculate share-of-model from a Prompt Panel?

    For each prompt, note whether your brand is cited or mentioned. Share-of-model = (prompts where you appear ÷ total prompts) × 100, per engine. Then average across engines for a portfolio score. Track the number monthly and tie deltas to specific placements or content ships.

    If You're Invisible in AI, You're Losing Clients Right Now.

    See exactly how your company appears across AI, search, and investor research — and uncover the hidden gaps costing you trust and deals.