> ## Documentation Index > Fetch the complete documentation index at: https://docs.get3rd.com/llms.txt > Use this file to discover all available pages before exploring further. # Topics, markets & audiences > Organize prompts by topic, geography, language, and audience so you can slice AI-visibility results, compare segments, and act on the right gaps. Before adding prompts at scale, set up the structure you'll filter and report by. Every metric in 3RD can be filtered by these three dimensions. ## Topics A **topic** is a named group of related prompts — "pricing", "integrations", "alternatives", "employer branding". Topics: * Organize your tracking by subject area * Unlock better prompt suggestions * Become a primary filter on the Leaderboard and every metric Think about the different areas where you want to *be the answer*, and create a topic for each. ## Markets A **market** is a **geography + language** pairing — *USA – English*, *Denmark – Danish*. They're tracked separately because AI answers differ by both dimensions: *Denmark – Danish* and *Denmark – English* are two different markets and can produce genuinely different answers. Add one market per region/language combination you sell in. ## Audiences An **audience** is a user profile representing the type of person asking the prompts — an enterprise buyer, a solo founder, a developer. AI models personalize answers, so different profiles may get different recommendations. Audiences let you compare how answers change across buyer types. ## A clean structure pays off Every metric in 3RD — visibility, sentiment, position, rank, citation share — can be sliced by **topic × market × audience × model**. A clean structure set up front means every future question ("how do we do on pricing prompts in Germany?") is one filter away instead of a re-tagging project. ## Related * [Setting up your prompts](/setup/prompts) — what goes inside the structure * [Understanding the Overview & Leaderboard](/results/overview-and-leaderboard) — where the filters pay off