Step 1 — GET CONVERSATIONS
Click ‘Get Fresh Conversations’ to send your prompts to Perplexity, ChatGPT, Gemini and Claude. Choose all entities, a type, or a single entity from the dropdown. A PIN is required before any API call. Every response is captured, parsed and scored automatically in one go.
Step 2 — READ THE CONVERSATIONS
The Conversations tab shows every AI response in full. Click any prompt to see what each model said — and whether it mentioned, attributed, or recommended Shepherd Neame. Citation markers link to their sources, and a Sources section appears below responses that include references. Active filters are shown as a banner above the list.
Step 3 — CHECK YOUR SCORES
The Insights tab shows Brand Recognition, Recommendation Rate and Sentiment as animated dials tracking progress toward their goals. Hover the circled “i” next to any metric to understand what it measures. Key Findings are sorted with positives first and issues last — negative findings link directly to Fix It. The client name is configurable in Admin for white-label use.
Step 4 — FIX IT
Fix It shows exactly what’s holding each entity back and what to change — a schema update, a content addition, or a page to create.
Step 5 — PROMPTS & CATEGORIES
The Prompts tab groups prompts by entity with category labels: Discovery, Brand Story, Product, and Experience. Categories can be set manually or auto-inferred from keywords. Hit “Backfill categories” to categorise all uncategorised prompts at once and save to the database. Hover the “i” on the Category column header to see what each category means.
Step 6 — DEEP DIVE
Click any entity in Fix It for a full diagnostic. agentWOW crawls your website, reads your schema, and tells you precisely what AI can and can’t see — and why.
METHODOLOGY
Full calculation framework: six dimension formulae (MR, AC, RS, AT, CP, BA), composite score models (base and weighted), parity (A.3), cross-platform divergence (A.4), trend (A.5), seasonal adjustment (A.6). Appendix A complete.