How visible is your brand to AI?
BrandLens shows how AI models mention, describe, compare and recommend your brand — with evidence attached to every claim, not a guess dressed up as a metric.
The visibility problem
Search visibility no longer tells the whole story.
Search once meant rankings and keywords — a page you could open and check. Brands are now discovered, described and recommended inside AI-generated answers instead, often with no visible signal that anything has shifted.
Traditional search
- Ranked pages
- Fixed positions
- Keywords
- Clicks to a page
AI discovery
- Generated answers
- Mentions in context
- Descriptions & comparisons
- Recommendations, with sources
BrandLens measures the second column — not the first.
How it works
Measured, compared, tracked — the same way, every time.
- 01
Observe
Category-relevant prompts are sampled across multiple AI models on a recurring schedule.
- 02
Compare
Your brand's mentions are set against named competitors on the same prompts, the same models, the same window.
- 03
Interpret
Patterns become findings — evidenced statements about visibility and recommendation, not raw counts.
- 04
Track
Findings are dated and repeated, so a real change is distinguishable from normal variation.
Intelligence demonstration
A window into how BrandLens thinks.
AI Mention Share
+2 pts vs. 20% category averageFavorable
Last 30 days · 4 models · 1,200+ prompts sampled
Illustrative example — fictional data
AI mention share is now above the category average for the first time, driven by a 4-point gain this quarter.
High evidence — measured across 4 AI models, 1,200+ prompts, 30-day window.
Visibility falls significantly when a prompt actively compares alternatives — exactly where a buying decision is being made.
Recommendation
Publish evidence-led comparison content addressing the highest-value comparison prompts.
Could recover an estimated 6–8 points of mention share in comparison-style prompts specifically.
Medium evidence — modeled from comparable content launches, not yet measured for this case.
Source: BrandLens
Competitive intelligence
A comparison built from evidence, not self-report.
Every competitor is sampled under the same prompts, the same models, the same window as the subject brand — no arbitrary weighting, no invented score.
Category average — 20%
Illustrative example — fictional data, for demonstration only.
Methodology & evidence
What's measured, what's inferred, and what isn't claimed.
BrandLens samples a defined set of category-relevant prompts across multiple leading AI models on a recurring schedule — a mention is counted when a brand, or a close, unambiguous variant of it, appears in a model's response. Some conclusions are measured directly from these samples; others are estimated, inferred from patterns across comparable cases rather than observed directly. BrandLens always states which is which.
Coverage is limited to the models and prompt set actually sampled. A finding never implies comprehensive coverage of every AI system in use, and confidence is stated at the level the evidence actually supports — never higher.
Large sample, direct measurement, multiple corroborating sources.
Smaller sample, early signal, or a single corroborating source.
Model-derived inference or extrapolation from comparable cases.
Connection to Orentys
The signal is detected. Understanding why it matters is next.
A visibility finding says what's happening — AudienceLab explains why it matters, and ContentFlow turns that into prioritized action.
Detect
BrandLens
What do you need to see more clearly?
If your team can't yet answer how AI describes, compares or recommends your brand, that's the starting point.