<- The AI Visibility Index
DTC: running shoes
As of 2026-06-07. 5 brands across 3 surfaces. Observed live by querying the assistants directly and recording their real answers. A reading on a date, not a permanent fact.
ChatGPT (OpenAI Ads) LIVE(3/3 queries)Google Gemini (assistant app) RUMORED(1/3 queries)Microsoft Copilot Ads LIVE(3/3 queries)
The status on each surface is its advertising status (LIVE = ads buyable today; ANNOUNCED / RUMORED = monetization coming or reported), not the reliability of the observation. Every cell was read live regardless. Coverage is not a full grid: some surfaces were read on fewer queries, so per-brand denominators differ.
| # | Brand | Cited | Cited rate | Avg position | Points | |
|---|---|---|---|---|---|---|
| 1 | Nike | 7/7 | 100% | 1 | 10.00 | This is you? See why -> |
| 2 | ASICS | 5/7 | 71% | 2 | 7.67 | This is you? See why -> |
| 3 | Adidas | 6/7 | 86% | 2.5 | 6.83 | This is you? See why -> |
| 4 | Brooks | 3/7 | 43% | 1.67 | 5.33 | This is you? See why -> |
| 5 | Hoka | 4/7 | 57% | 3 | 4.75 | This is you? See why -> |
Not enough data to rank (cited too few times across this sample to place fairly, listed not scored): Saucony, New Balance, On Running
Head queries read
best marathon running shoesbest running shoes for beginnersbest running shoes
Observed live: every cell is grounded in the real answer an assistant gave to a real query, captured directly in the browser on the date shown, not inferred from the public corpus. A reading on a date, not a permanent fact. Coverage is not always a full grid: a surface may be read on fewer queries than another, so per-brand denominators can differ (see the per-surface query counts). The LIVE / ANNOUNCED / RUMORED label on each surface is its ADVERTISING status (whether ads are buyable there), not the reliability of the observation; every cell was read live regardless. A brand named without an explicit rank is counted as cited (one point, no position bonus) and shows average position n/a. Some assistants personalize to the signed-in user, so a reading is one segment, not a universal ranking. Brands with too few observed citations are listed as insufficient-data, not ranked. Where a labeled sponsored slot was shown it is recorded against the observed advertiser.
Sources
Are you on this list, or missing from it? Run your own AI Visibility Audit.