Brand Climb
AI assistants now send buyers to brands. Are they sending them to you or to a competitor?Check my score
Research

How we measure, and what we're learning

Here's how the numbers are produced, and what we can and cannot say from real data today.

Method, in the open

The score is a formula you can check

We publish the full AI Visibility Score formula and what each input means, so you can check any number we show.

  • Every metric traces back to the stored AI answer it came from
  • Sentiment is read only from sentences that name your brand, and the method (keyword-based) is disclosed
  • Competitor sub-metrics we can't measure are labelled as estimates, not passed off as measured
  • Demo data is always labelled as demo data
See the full formula
Not mentioned

“For performance under $3,000 most riders shortlist the Trailforge Apex and the VeloPeak R2. Both publish full geometry…”

What we're publishing next

Findings

We will publish aggregate findings, such as how visibility differs across engines, which source types earn the most citations, and how answers change over time, once we have enough real runs for them to be reliable. Until then this section stays empty. When we have findings, they'll appear here and on the blog.

Run the method on your own brand.

The free report applies the same measurement to your brand.