← Back to Blog

October 16, 2026

The Nike Paradox: AI talks about Nike more than anyone, and likes it less than almost anyone

Ask an AI assistant anything about running shoes and there's a very good chance the word "Nike" comes back in the answer.

We measured exactly how good. Then we found something more interesting.

This week we ran an EYE study on Nike: 50 real consumer questions about running shoes and sportswear, asked across 8 AI platforms (Google, OpenAI, DeepSeek, Grok, Anthropic, Perplexity, Llama and Kimi), in US English, on 16 October 2026. We read every answer and counted every brand, every source, every adjective.

(Nike is not an EYE client. This is an independent study of a public brand, shared because it's one of the clearest examples we've seen of how AI search really works.)

Here's what we found, and what it means for any brand, Nike-sized or not.

1. Nike wins the volume battle by a landslide

  • 58.4% share of voice. More than every other brand combined.
  • 3,171 mentions, compared with adidas at 380 and ASICS at 366. That's more than 8 to 1.
  • The word "nike" appeared 945 times in the answers. "Shoes" appeared 478 times. In AI's vocabulary, Nike is used more often than the product category itself.
  • GSEO score: 91 out of 100. Excellent visibility, high answer quality, stable trend.

By any traditional measure of brand presence, this is total dominance. If AI search were a market, Nike would be a monopoly.

But volume is only half the story.

2. The paradox: the most mentioned brand is one of the least loved

When we scored how AI describes each brand (from -1, negative, to +1, positive), the ranking flipped:

  • On: 0.82
  • Saucony: 0.71
  • Hoka: ~0.68
  • ASICS: 0.63
  • Brooks: 0.59
  • New Balance: 0.57
  • Puma: 0.50
  • adidas: 0.39
  • Nike: 0.33

Nike is mentioned 36 times more often than On, but described less warmly than every brand in its competitive set.

Let's be fair: Nike's sentiment isn't bad. Of the mentions with a clear tone, 546 were positive and only 23 were negative. That's 2% negative. The problem isn't criticism. The problem is the 431 neutral mentions. AI talks about Nike the way you talk about the weather: it's always there, it's rarely exciting.

The challengers are mentioned less, but when they are, AI recommends them with enthusiasm. "If you want maximum cushioning, Hoka." "For a lightweight, eco-conscious option, On." Specific reasons. Warm words.

Why this matters: a buyer asking AI for a recommendation doesn't count mentions. They read the sentence. And the sentence that says "best for your needs" is worth more than ten that say "Nike also makes running shoes."

3. Nike is losing the questions that sound like buying decisions

Look at which brand AI placed first on each question, and a pattern appears:

  • "How durable are Nike running shoes compared to Adidas?" Top brand: adidas. Nike lost a question with its own name in it.
  • "What types of shoes are best for trail running?" Top brand: Hoka.
  • "Are there eco-friendly running shoe options besides Nike?" Top brand: On.
  • "What are some reliable running shoe brands besides Nike and Adidas?" Top brand: Saucony.
  • "Can you recommend running shoes similar to Nike but cheaper?" Top brand: Puma. This question generated 171 brand mentions, one of the busiest in the study.

And where does Nike win? On questions like:

  • "How can I be sure I'm buying authentic Nike shoes?" (position 1)
  • "How can I tell if a Nike store is an authorized retailer?"
  • "Why do some Nike running shoes cause discomfort?"
  • "What makes Nike sportswear stand out?"

Read those two lists again. Nike dominates the questions people ask about Nike. Challengers win the questions people ask when they're still deciding.

That's the most important insight in the whole study, and it applies to almost every category leader we track.

4. The content gaps are specific, and fixable

EYE measures the gap between competitor mentions and yours on each topic. Nike's biggest:

  • Best support for long-distance running: competitors 40 mentions, Nike 13. Gap of 27.
  • Eco-friendly running shoes: competitors 18, Nike 5.
  • Trail running: competitors 15, Nike 3.
  • Arch support (Nike vs ASICS): competitors 60, Nike 49.

None of these are mysteries. Nike makes trail shoes. Nike makes stability shoes. Nike has a sustainability programme. AI just doesn't connect Nike to these needs, because the content that would make the connection either doesn't exist, isn't specific enough, or doesn't live where AI looks.

Which brings us to sources.

5. Nike is its own best source, but not its only one

The good news for Nike: nike.com is the most cited domain in the study (172 citations), plus 57 more from about.nike.com. When AI needs facts about Nike, it goes to Nike. Few brands achieve that.

But look at what comes next:

  • Runner's World (95 citations)
  • RunRepeat (73)
  • RTINGS (46)
  • Dick's Sporting Goods (33)
  • Then a long tail of review sites, running blogs, foot health sites, PubMed research and retailers, across 50 domains.

Notice who wins the opinion questions. Not the brand site. The reviewers. Nike.com tells AI what Nike makes. Runner's World and RunRepeat tell AI what's good. Sentiment is earned in third-party content, and that's where the challengers have been quietly investing for years: running specialty press, lab-style reviews, community recommendations.

One more detail: when AI described where to buy Nike, it listed nike.com and Foot Locker, but also StockX, GOAT, eBay, Poshmark, TJ Maxx and Ross. AI is building its own map of your distribution, including the channels you'd rather it didn't recommend.

6. Each AI sees a different Nike

Same 50 questions, very different results by platform (Nike mentions per answer):

  • Google: 33.2
  • OpenAI: 23.1
  • DeepSeek: 23.1
  • Grok: 15.8
  • Anthropic: 10.5
  • Llama: 1.6

Google mentions Nike three times as often per answer as Anthropic's Claude, and 20 times as often as Llama. Answer quality and completeness also varied widely between platforms.

If your team "checks ChatGPT" once a month and calls it AI monitoring, you're seeing one of these numbers. Your customers are using all of them.

What this means for your brand (even if you're not Nike)

1. Share of voice is not the finish line. You can own the conversation and still lose the recommendation. Track sentiment and position alongside mentions, or you'll be measuring the wrong thing.

2. Map your questions by intent. Split the questions in your category into "about us" and "deciding". Most brands win the first group and lose the second without knowing it. Nike's study had 28 informational, 9 comparison, 7 problem-solving and 6 transactional questions. The comparison and "besides X" questions are where challengers steal the sale.

3. Every "besides [leader]" question is an open door. If you're a challenger, these are the questions to own. If you're the leader, these are the questions to defend, by giving AI clear, specific reasons you're still the right answer.

4. Write for the need, not the product. "Best shoe for long-distance support" will be answered by whoever writes the clearest, most specific, best-evidenced content about long-distance support. Product pages with taglines won't do it.

5. Earn the third-party sentence. AI forms opinions from reviewers, specialists and communities. Getting your product properly reviewed by the 5 to 10 sources AI cites in your category may be worth more than any ad campaign this year.

6. Measure across platforms, every week. AI answers change. Platforms disagree. A single snapshot is an anecdote; a weekly trend is a strategy.

The bigger picture

Nike has the most powerful brand in sport, a world-class website, and more mentions than its next eight competitors combined. And AI still sends the "what should I buy?" moments to challengers with better-described products.

If that can happen to Nike, it's almost certainly happening in your category. The only question is whether you can see it.

That's why we built EYE: to show brands exactly what AI says about them across 14 AI platforms, 203 countries and 61 languages, who it recommends instead, which sources decide, and the ranked list of fixes that change the answer.

See what AI says about your brand: https://eye.synapsedatagroup.com/?utm_source=linkedin&utm_medium=article&utm_campaign=nike_paradox

Methodology: EYE AI Visibility Review, 50 consumer questions on running shoes and sportswear, 8 AI platforms, United States, English, run on 16 October 2026. Sentiment scored -1 to +1 per brand mention. Perplexity returned 6 of 50 responses and Kimi 5 in this run, so their figures are excluded from the platform comparison. Nike is not affiliated with EYE or Synapse Data Group.

#GEO #AISearch #Nike #BrandStrategy #RunningShoes #ChatGPT #Marketing #CMO

Promo post to share the article (post this the same day)

Nike owns 58% of the AI conversation about running shoes.

It's also described less warmly by AI than On, Saucony, Hoka, ASICS, Brooks, New Balance, Puma and adidas.

We asked 8 AI platforms 50 questions. Nike won almost every question with "Nike" in it.

And lost the ones that sound like buying decisions: → Trail running? Hoka. → Eco-friendly? On. → Similar but cheaper? Puma. → Nike vs Adidas durability? adidas.

#GEO #AISearch #Nike #Marketing