Brand Mention

When an assistant answers "best affordable fragrances" and writes your brand into the response, that's a mention. It sounds trivial, but detecting it reliably is not: models use full names, initialisms, and product names interchangeably, and the same words can refer to different entities. Good mention detection has to handle all of that.

• Foundation for every metric. Share of Answer, prominence, and competitive benchmarks all build on an accurate mention count.
• The first bar to clear. Being mentioned is the minimum viable outcome of GEO: you can't be prominent in an answer you're absent from.
• A leading indicator. Rising mention frequency across your prompts is early evidence that models are learning to include you.

Detection normalizes the answer text, flattening case, accents, and punctuation, and matches it against each entity's name and aliases. That normalization is why "Carolina Herrera," "carolina herrera," and a registered alias all resolve to the same brand, and why maintaining aliases is essential to catching every legitimate mention while avoiding false matches on ambiguous terms.

A mention is your brand named in the prose. A citation is your source linked as grounding for the answer. They're related but distinct: you can be mentioned without being cited, and occasionally cited without being named prominently. Strong visibility usually means both: named in the answer and cited as a source.

Earn broad, consistent presence across the web so models associate your brand with the category; publish citable content that answers real questions; and keep naming and aliases consistent so every reference is captured.

Is one mention per answer the maximum?

No. A brand can be referenced multiple times in one answer. Counting and prominence weighting handle repeated and emphatic mentions.

What's the difference between a mention and prominence?

A mention is binary (named or not). Prominence weighs how strongly you're named: position, emphasis, and whether you're cited.