A strong topic structure does more than keep your dashboard tidy. It shapes how you interpret performance, communicate results to stakeholders, and decide where to invest in content next.

The most useful way to approach topics isn't "what should I call these folders?", it's "what are the competitive areas that actually matter to my business?"

Your topic structure should map to the markets, audiences, and intent clusters where your brand needs to show up in AI-generated answers. Before start creating topics, ask yourself:

The answers to these questions are your topic candidates.

💡 Keep in mind: Topics you create now anchor your historical data. A little upfront thought pays off, but you can make updates to your initial setup.

Your topic structure should reflect how your team talks about performance. If your team reports by product line, topics should map to products. If they report by customer segment, topics should map to segments. If you're running a campaign and need to measure its AI search impact, create a topic for it.

A topic called "Salon Software - BOFU" tells a story: it's the bottom-of-funnel visibility for your salon software product. When visibility climbs in that topic after you publish a comparison guide, you know exactly what moved the needle and you can report it clearly.

ICP-based topics work best for B2B companies with distinct customer segments that search differently. Each topic covers a specific buyer type, letting you correlate content published for a specific ICP with visibility changes in that topic.

Example: Security Compliance - Growth Stage Saas, Security Compliance - Enterprise, Security Compliance - Startup Saas

Funnel-stage topics work best for teams that want to understand where in the decision journey their brand is visible. Research-intent prompts ("what is salon scheduling software?") behave very differently from purchase-intent prompts ("best salon booking software for a chain").
​
Splitting them into TOFU, MOFU, and BOFU topics lets you track both, and see whether you're winning where purchase decisions get made.
​

Geography-based topics work best for service businesses or companies where location is a meaningful filter for buyers.

Example: Warehouse Fire Protection - Midwest, Warehouse Fire Protection - Southeast.

These aren't mutually exclusive. As your tracking matures, you may layer more than one approach.

Topics work best when each one has a clear, distinct purpose. A useful test: if you can immediately describe what type of prompt belongs in a topic (and what doesn't), it's well-defined. If you'd debate which topic a new prompt belongs to, the structure might need tightening.

⚠️ Caution: Deleting a topic deactivates all prompts within the topic.

The people closest to your customers — sales, customer success, product — often have the best instincts for how buyers talk about your category. Before finalizing your topic structure, ask them: what questions come up most in sales calls? What objections do buyers raise? What comparisons do they make? Their answers can surface topic categories and prompt ideas that data alone won't give you.

Topics aren't set-and-forget. Review your topics at the start of each quarter. For each one, ask whether it still maps to an active business objective and whether the prompts inside it still reflect how buyers are asking questions. Deactivate topics that are no longer relevant rather than deleting them, so you preserve the historical data they contain.

Once you have a plan, the Step-by-Step Guide: Creating & Managing Topics walks you through every platform action, from creating your first topic to deactivating ones you no longer need.