The 73% / 44% / 11% ladder: whether AI mentions you starts with brand stature
Same prompts, first run: global brands appear in 73% of relevant answers, mid-market and regional brands 44%, small and niche brands 11%.
Key numbers
- Sample: 100+ brands and 100,000+ prompt responses collected March–May 2026.
- First-run visibility forms a three-tier ladder: global household names 73%, established mid-market and regional brands 44%, niche and small brands 11% — roughly 30 percentage points per step.
- About 78% of cited sources point to corporate websites; among non-corporate sources YouTube leads, ahead of Reddit, editorial media and Wikipedia.
- The most-cited content format is the ranked “best-of” listicle at about 21% of citations. Sentiment is the unstable signal: how positively or negatively a brand is framed flips about 6.7× more often than whether it is mentioned at all.
Method and sample
arXiv paper “Generative Engine Optimization at Scale: Measuring Brand Visibility Across AI Search Engines” (2026-06-18): analysis of 100+ brands and 100,000+ prompt responses from March to May 2026, measuring first-run visibility, source composition, cited content formats and sentiment volatility.
What it means for brands
Brand stature converts directly into AI exposure, and the ladder is not linear: small brands are not mentioned slightly less, they are close to absent. And once you are mentioned, how you are described flips far more often than whether you appear — so track it per snapshot.
How SourceSight handles it
SourceSight records mention rate and position per keyword per snapshot and compares consecutive runs, so you can tell whether what moved is appearance or description; prompt libraries are grouped by how customers actually ask, so the entirely missing question types get fixed first.
Limits
The sample comes from a commercial monitoring panel and skews to Western markets and English prompts; the ladder is an observed correlation, not a guarantee that any single action moves a brand a tier.
References
Run it on your own project
What the same prompts return across 10 engines is only knowable by running them.
Related research
What 2 million AI citations show: prompt-content match matters most, checklists matter less
Once you control for the domain, the usual AEO checklist (FAQ blocks, structured data, Core Web Vitals) reverses or collapses to zero — domain authority and prompt alignment dominate.
How engines work · 2026-10-07Whose side is the AI shopping assistant on? Change its role and the penalty for sponsored results disappears
The same sponsored listings get treated very differently once the prompt names the platform as the agent's principal — disclosure rules built for humans do not carry over.