Methodology

How we measure the layer of visibility AI now controls.

We evaluate the structure behind recommendation outcomes — brand presence, citation frequency, source influence, competitive share, and the patterns that determine when a brand is surfaced or left out.

Framework

Four layers of analysis.

Prompt set analysis

We evaluate commercially relevant prompts and category-specific questions designed to reflect how buyers discover, compare, and assess brands through AI-generated answers.

Response and citation extraction

We analyze model outputs to identify which brands appear, which sources are cited, and how authority is constructed within the response.

Competitive benchmarking

We compare brand visibility against relevant competitors to understand share of presence, share of citation, and relative recommendation performance across the same query set.

Source influence mapping

We identify the domains, publishers, and referenced materials most associated with visibility and recommendation outcomes, revealing the source architecture shaping model trust.

Five dimensions

What we measure.

Brand presence

How often a brand appears in relevant AI-generated responses.

Citation visibility

The extent to which a brand is supported by cited or source-linked information within model outputs.

Competitive share

A brand's relative visibility compared with peer brands across the same prompt environment.

Source influence

The degree to which specific sources contribute to brand authority and recommendation patterns.

Recommendation positioning

How a brand is framed in AI answers — surfaced positively, comparatively, or not at all.

Interpretation

How to read the data.

Repeatable patterns over single answers

AI outputs are probabilistic. Our methodology measures durable signals across structured analysis rather than overstating any single response.

Where authority is accumulating

Citation patterns reveal which sources are increasingly trusted by AI systems — a leading indicator of recommendation strength.

Where competitors are outperforming

Benchmarking exposes the prompt clusters and citation networks where the gap is widest and most recoverable.

Methodology built for boardroom decisions.

Brand performance increasingly depends on whether AI systems recognize a company as credible, relevant, and worth recommending. Our methodology is built to measure exactly that.