Topic: A container that groups related prompts around a coherent area of intent. Topic-level metrics aggregate from the prompts inside, so mixing unrelated prompts dilutes your data.
Coverage topic: A broad topic that monitors whether you're in the conversation for a category (e.g., "historical data"). Answers: are we visible here?
Depth topic: A focused topic designed to win a specific opportunity (e.g., "historical data for marketers"). Answers: are we the default recommendation?
Commercial intent prompt: A prompt that forces the answer engine to recommend rather than explain. Outputs typically include "best tools," "top platforms," or "recommended options."
Query fan-out: When a user asks a complex question, answer engines break it into multiple sub-queries, fetch sources, then synthesize. Winning specific sub-queries helps you appear in broader answers.
Keep both in your setup:
Start with a base (your product/category term), then add qualifiers:
Compound prompt example: "What's the best AI search platform for marketing teams who need historical visibility data to justify their 2026 budget?"
Avoid brand names in prompts as they skew visibility scores.
Group prompts where:
If competitors and criteria differ significantly, split into separate topics.
Topic: A container that groups related prompts around a coherent area of intent. Topic-level metrics aggregate from the prompts inside, so mixing unrelated prompts dilutes your data.
Coverage topic: A broad topic that monitors whether you're in the conversation for a category (e.g., "historical data"). Answers: are we visible here?
Depth topic: A focused topic designed to win a specific opportunity (e.g., "historical data for marketers"). Answers: are we the default recommendation?
Commercial intent prompt: A prompt that forces the answer engine to recommend rather than explain. Outputs typically include "best tools," "top platforms," or "recommended options."
Query fan-out: When a user asks a complex question, answer engines break it into multiple sub-queries, fetch sources, then synthesize. Winning specific sub-queries helps you appear in broader answers.
Keep both in your setup:
Start with a base (your product/category term), then add qualifiers:
Compound prompt example: "What's the best AI search platform for marketing teams who need historical visibility data to justify their 2026 budget?"
Avoid brand names in prompts as they skew visibility scores.
Group prompts where:
If competitors and criteria differ significantly, split into separate topics.
