A first look at AI search visibility for the world’s leading athletic apparel brands. May 2026 data.
Inside ChatGPT, one of the most recognized brands in athletic apparel is in free fall. Puma has lost 23 points of brand visibility in 110 days. The steepest decline in AiRR’s seven-industry dataset.
Here is what produced that finding. 110 days of continuous tracking. 1,829 data points pulled from ChatGPT. Six brands. Three buyer personas. Two purchase-intent queries about athletic footwear. This is the May 2026 AiRR snapshot for athletic apparel.
You cannot get this data by opening ChatGPT yourself. A single user gets a single personalized answer at a single moment in time. AiRR runs structured queries at scale, in mass quantities, across multiple personas, every single day. That is how the AiRR Score is built: a 0 to 100 composite that measures every response across four dimensions: Presence (does the brand appear?), Perception (how is it described?), Prestige (is it recommended above competitors?), and Persistence (does visibility hold over time?).
Six tracked athletic apparel brands ranked by AiRR Score on ChatGPT, May 2026.
We Tracked 6 Top Brands
This snapshot measures how ChatGPT recommends athletic apparel brands in response to two purchase-intent queries: “Give me a top 10 of the best outdoor basketball shoe brands” and “Give me a top 10 of the best outdoor track shoe brands.” Both queries run at scale, in mass quantities, with no individual user signals attached.
AiRR tracks six brands in athletic apparel: Nike, Adidas, Under Armour, Asics, Reebok, and Puma. Other brands appear in ChatGPT’s responses; the AiRR Score below reflects how these six perform across the full set of responses.
The leaderboard for May 2026:
- 1. Nike 89.50
- 2. Adidas 81.30
- 3. Under Armour 70.68
- 4. Asics 60.73
- 5. Reebok 55.47
- 6. Puma 51.93
Nike’s lead at the top is durable. The brand scores nearly 90 out of 100, dominating Presence (97.77) and Persistence (97.54). What ChatGPT calls the most reliably-recommended athletic apparel brand for the broadest range of buyer queries is the brand most people would have guessed. The category leader leads.
Adidas at 81.30 is closer than the gap looks. Where Nike dominates Presence, Adidas is more competitive on Perception (81.39 versus Nike’s 82.38), suggesting that when ChatGPT does mention Adidas, it describes the brand as favorably as it describes Nike. The opportunity for Adidas is closing the Presence gap, not changing the conversation.
Under Armour at 70.68 holds third. Asics, Reebok, and Puma trail at fourth, fifth, and sixth.
The Real Story Is the Movement
Static rankings hide the most important pattern. Brand visibility inside ChatGPT moves faster than traditional brand metrics suggest.
AiRR Score over time, January 23 to May 13, 2026. Live tracking from the AiRR platform.
Across the 110-day window:
- Puma lost 22.85 points of AiRR Score. The steepest decline in the entire seven-industry AiRR dataset.
- Asics gained 8.05 points.
- Adidas lost 6.20.
- Reebok lost 5.45.
- Nike lost 2.08 (essentially flat).
- Under Armour lost 0.75 (essentially flat).
Puma’s 23-point decline is the kind of brand signal that public-market analysts will eventually price in.
Asics is the quiet winner. Up 8 points in 110 days, the brand is closing on Under Armour’s third position from below. ChatGPT is increasingly recommending Asics in the prompts where it matters.
“The brand that wins the AI recommendation depends on who is asking. Optimizing for general AI visibility is not the same as optimizing for the customer who actually buys.”
The Hidden Variable Is Persona
The AiRR Score is the headline number. The deeper insight is that the score changes substantially depending on which audience persona ChatGPT is responding to.
When ChatGPT answers for an athletic buyer, Reebok drops 33 points, Puma drops 27, and Asics gains 9.
When AiRR runs queries with an Athletic and Active persona instead of an Overall persona, the rankings change:
- Reebok drops 33 points (55.47 to 22.47). The brand has recognition with general audiences but loses the recommendation when the buyer knows the category.
- Puma drops 27 points (51.93 to 24.63). Same pattern.
- Asics moves the opposite direction. Up 9 points (60.73 to 69.92). Informed buyers recognize Asics as a serious performance brand.
- Nike and Adidas are stable across personas. Under Armour drops moderately.
This is the persona reversal effect. The brand that wins the recommendation depends on who is asking. For athletic apparel CMOs, the implication is direct: optimizing for general AI visibility is not the same as optimizing for the actual customer persona that drives purchase.
What this means for athletic apparel marketers in May 2026
- If you are Nike or Adidas, you are winning the broad consumer query. Hold the position. The risk is complacency, not competition.
- If you are Under Armour or Asics, your opportunity is in the persona-specific prompts where the field is more open. Asics gained 9 points with athletic buyers. There is room to compete on category-specific queries that the giants do not own outright.
- If you are Reebok or Puma, the brand recognition you have with general consumers is not converting inside informed-buyer prompts. Optimize for category-knowledgeable personas first. The general audience will follow.
- If you are an investor or analyst covering this category, AiRR Score declines often precede traditional brand-tracking declines by months. Puma’s 23-point drop is the kind of leading indicator that public markets will eventually price in.
What’s your AiRR Score? Find out at airrscore.com.
Methodology
AiRR is the independent measurement platform for brand visibility in AI search. The methodology runs structured queries to ChatGPT at scale, in mass quantities, across multiple buyer personas, every single day.
This is not something an individual can replicate by opening ChatGPT and typing the same questions. When you query ChatGPT directly, the model already knows who you are. Cookies, account memory, prior conversations, location signals, and dozens of other inputs shape the response before you finish typing. There is no “blank slate” at the consumer level. Every individual answer is already a personalized answer.
AiRR’s queries run independently of those signals. Every brand is measured against the same baseline. Every score is reproducible. Every comparison across competitors and across time is consistent. That is the data layer that no other tool in the GEO and AEO category provides.
The four dimensions of the AiRR Score, visualized for the six leading athletic apparel brands. Nike covers the largest area; Reebok and Puma are constrained on Prestige.
The 4P framework breaks down brand visibility into Presence (does the brand appear?), Perception (how is it described?), Prestige (is it recommended above competitors?), and Persistence (does visibility hold over time?). Each dimension is scored 0 to 100. The composite AiRR Score weighs all four.
This snapshot is built from 837 brand-level data points and 992 prompt-level data points, collected between January 23 and May 13, 2026.
About the author
Steven Perlman built AiRR (airrscore.com) because AI search needed a measurement layer that did not also sell the cure. AiRR scores brand visibility in ChatGPT and other AI platforms. He also founded Syfter, a New York technology staffing firm.
Press inquiries: steven@airrscore.com
