Response Coverage

Response Coverage assesses how thoroughly AI engines represent a brand when they do mention it. A brand might have strong Share of Model (appearing frequently) but poor Response Coverage (being described in a limited or outdated way). This gap between mention frequency and mention quality is what Response Coverage exposes. The metric provides the qualitative depth dimension that pure frequency metrics miss.

The metric works by comparing AI-generated brand descriptions against a canonical set of brand attributes: core products, key features, differentiators, pricing model, target audience, and value propositions. If your brand offers five products but AI engines consistently mention only one, your Response Coverage is incomplete. If your key differentiator is security compliance but AI engines never mention it, that gap represents a specific optimisation opportunity that can be addressed through targeted content and structured data.

Response Coverage problems typically have identifiable root causes. Incomplete coverage often stems from: thin or disorganised website content that fails to clearly communicate all brand attributes, training data that reflects an older version of the brand, competing information from third-party sites that presents an incomplete picture, or missing structured data (llms.txt, llm-profile.json) that would provide AI engines with a comprehensive brand summary.

Improving Response Coverage requires a systematic approach. First, define the canonical set of brand attributes that should appear in AI responses. Then audit current AI responses to identify gaps. Finally, address each gap through targeted content, structured data, and citation network improvements. The goal is not just to be mentioned but to be mentioned completely and accurately.

The llms.txt file and llm-profile.json play a particularly important role in Response Coverage because they provide AI engines with a canonical, structured summary of the brand's full offering. Without these files, AI engines must piece together brand attributes from scattered web content, making incomplete coverage far more likely. Implementing these files is often the single highest-impact action for improving Response Coverage.