We ran 800+ buyer prompts across ChatGPT, Perplexity, and Claude for 175 ecommerce brands in UK, US, and AU. Here's who AI actually recommends — and who's invisible.
| # | Brand | Category | Region | Score | Bar |
|---|---|---|---|---|---|
| 1 | beauty | US | 91/100 | ||
| 2 | beauty | US | 88/100 | ||
| 3 | beauty | UK | 84/100 | ||
| 4 | beauty | US | 81/100 | ||
| 5 | beauty | US | 76/100 | ||
| 6 | beauty | US | 72/100 | ||
| 7 | beauty | AU | 72/100 | ||
| 8 | beauty | UK | 72/100 | ||
| 9 | beauty | UK | 71/100 | ||
| 10 | beauty | UK | 69/100 | ||
| 11 | beauty | UK | 69/100 | ||
| 12 | beauty | US | 67/100 | ||
| 13 | beauty | UK | 67/100 | ||
| 14 | beauty | UK | 62/100 | ||
| 15 | beauty | UK | 62/100 | ||
| 16 | beauty | UK | 60/100 | ||
| 17 | beauty | US | 58/100 | ||
| 18 | beauty | AU | 58/100 | ||
| 19 | beauty | UK | 58/100 | ||
| 20 | beauty | UK | 57/100 | ||
| 21 | beauty | UK | 56/100 | ||
| 22 | beauty | US | 55/100 | ||
| 23 | beauty | UK | 53/100 | ||
| 24 | beauty | UK | 53/100 | ||
| 25 | beauty | UK | 52/100 | ||
| 26 | beauty | UK | 51/100 | ||
| 27 | beauty | UK | 49/100 | ||
| 28 | beauty | AU | 48/100 | ||
| 29 | beauty | UK | 47/100 | ||
| 30 | beauty | UK | 46/100 | ||
| 31 | beauty | UK | 44/100 | ||
| 32 | beauty | UK | 43/100 | ||
| 33 | beauty | UK | 42/100 | ||
| 34 | beauty | UK | 41/100 | ||
| 35 | beauty | US | 40/100 | ||
| 36 | beauty | UK | 39/100 | ||
| 37 | beauty | UK | 38/100 | ||
| 38 | beauty | UK | 29/100 |
Enter your store URL to get your free AI Visibility Score — the same methodology used to rank all 175 brands in this report.
Free · Takes 2 minutes · No credit card required
Every brand was scored using a battery of 8 buyer-intent prompts per category — prompts like "best [category] brands in 2026", "recommend me a [category] brand that ships to [region]", and "what's the top [category] for [use-case]".
Prompts were run across three AI platforms: ChatGPT (GPT-4o), Perplexity, and Claude 3.5. Each mention = points. Ranking higher in a recommendation = more points.
Scores are normalised to a 0–100 composite. 60+ = strong presence. 40–59 = visible but patchy. Below 40 = significant gaps.
This report is designed to be citation-credible — the methodology is deterministic, repeatable, and platform-agnostic. You can replicate any result by running the same prompts yourself.