Glossary
Branded share
The proportion of prompts in a batch that explicitly mention your brand, as opposed to those that stay generic ("what's the best portable AC?" vs "is portable AC X any good?"). Expressed as a percentage of the batch.
Concrete example
Out of 100 prompts, 30 name "X" ("X vs Y comparison", "X reviews"), 70 stay generic. Branded share = 30%.
Why it matters
Too much branded (>60%) biases the measure toward answers that necessarily cite you. Too little (<10%) makes it hard to test recommendation scenarios. The right mix depends on your notoriety.
Related terms
- Core vs Rotation
Two categories of elements in a prompt batch. Core (brand, fixed personas) stays constant across the batch to keep a comparable baseline. Rotation (topics, intents) varies prompt by prompt to cover a spectrum.
- GEO (Generative Engine Optimization)
The practice of optimising a brand's presence and framing in generative engine answers (ChatGPT, Claude, Gemini, Perplexity). GEO complements SEO: SEO targets blue-link rankings, GEO targets inclusion in the synthesised answer.
- Credit
The billing unit debited for each prompt generated via an LLM. One prompt = one credit at generation time. Prompts marked "Very close" (near-duplicates) or "Auto fallback" (technical AI-call failure) do not consume credit on deletion.