If you were training for a marathon, you’d probably start with a new pair of running shoes. But if you asked ChatGPT which ones to buy, would its recommendation reflect what runners, independent testers and customers say?
Watch on YouTube
In this GEOMAX case study, Justin Hodgson, James Terry and Liam Hodgson compare Nike and HOKA across business performance, independent shoe testing, customer sentiment and AI recommendations.
Using 40 US-market questions, we explore which brands ChatGPT introduces, how well it understands them, which sources it cites and what happens when it must choose directly between Nike and HOKA.
The findings raise an important question for any business: does your real-world reputation carry through to the recommendations customers receive from AI?
Watch the video to discover the results and what they could mean for your business.
Read the full transcript
If you were going to run a marathon, you’d probably want a new pair of running shoes. But where would you start? Many people ask ChatGPT. So we asked ChatGPT which running shoes it would recommend between Nike and HOKA, two of the largest and fastest-growing brands in the world. This is a really interesting case study: which shoe ChatGPT recommends, why, and what that means for your business. I’m here with Liam and James, and we’re going to ask that question today and see what the results are.
When you analyse Nike versus HOKA in the physical world, you find some very interesting facts. Nike is 17 times bigger than HOKA, with annual brand revenue of about $45.2 billion. It’s a behemoth, renowned for cutting-edge technology and elite athlete endorsement. If you only looked at size, you might back Nike. It has decades of recognition, world-class athletes and a deep history in running.
HOKA, on the other hand, is much smaller, with annual brand revenue of about $2.6 billion. However, HOKA is growing 16 times faster than Nike. People love HOKA. It’s one of the fastest-growing shoe brands in the world. People recommend HOKA to their friends, and salespeople often recommend HOKA to customers in store.
HOKA has done something every challenger brand dreams about. For some customers it’s not just a shoe; it’s the shoe they insist their friends should try. That’s the reason its comfort story travels so well. Its thick cushioning attracts runners, people on their feet all day and customers seeking comfort. Word of mouth drives recommendations. Customers tell their friends, and salespeople may recommend HOKA when someone walks into a store. Word of mouth becomes a real-world recommendation engine.
So Nike wins on scale and heritage. HOKA wins on current growth and consumer recommendations. But neither figure tells us which shoe ChatGPT will tell a runner to actually buy.
So we investigated the technical side of the shoes. To do that, we checked published laboratory scores from RTINGS. These are model-level scores out of 10, not brand quality percentages. Marathon scores reflect suitability for long-distance running, cushioning and stability. The Nike Alphafly 4 has the strongest marathon rating in this group at 8.4. The Alphafly 3 has the highest stability score of the five shoes. HOKA’s Cielo X1 3.0 has the highest cushioning score at 9.1. Both HOKA racing models score 8 out of 10 for marathon racing. Both brands have strengths, despite the size gap.
Then we looked into brand sentiment: which shoes do marathon runners report liking more? We sourced these figures from RedditRecs, which analyses discussions on Reddit about specific running shoes. In that data set, the HOKA shoe was better liked than the Nike shoe.
Next, James looks at something most shoppers will never see: how clearly the brands’ websites present the evidence behind their products. We call this our technical SEO score.
Our GEOMAX technical SEO score measures how easily AI tools can access, read and make sense of the product information on a brand’s website. On that score, Nike gets 81 out of 100 and HOKA gets 53 out of 100, based on the product pages we examined.
When we dive into the numbers, both sites generally let AI systems access their pages and page content, and both put product information in readable page code. But Nike’s structured product details are definitely more complete and easier to find. Think of it as a neat digital product label: the model, the size, the colour, the price and where you can buy it. Nike also has a distinct advantage in that it has separate review pages with readable customer reviews.
Also surprisingly, neither brand publishes stack height, the thickness of the sole, on its own site. So independent reviewers are filling that gap for shoppers and AI tools. We won’t go through every fix today, but both brands have a really clear opportunity to help tools like ChatGPT recommend them.
So, in summary: Nike is much bigger. HOKA is growing much faster. One of Nike’s shoes is independently rated as the best overall marathon shoe, but one of HOKA’s shoes has the best cushioning. And Reddit users suggest they like one of the HOKA shoes better than the Nike. Nike, however, is doing a much better job of helping AI tools like ChatGPT find the content they need to make a decision.
All right, we’ve made the case for both brands. Liam, share the results with us.
We ran 40 US-market questions through ChatGPT using Ahrefs Brand Radar. Ten were open buying questions that named neither brand (Be Found). Ten tested brand understanding (Be Understood). Ten examined information and sources for the comparison (Be Trusted). And ten asked ChatGPT to choose directly between Nike and HOKA (Be Chosen).
Nike was not the only brand appearing regularly. Adidas, ASICS and Saucony also appeared in eight of the ten answers. HOKA was falling behind that established group. HOKA made the first four recommendations only once; Nike made the top four eight times. HOKA isn’t consistently getting into the initial shortlist. In the ten Be Found questions, Nike appeared in eight and HOKA in just two.
So maybe ChatGPT just doesn’t understand HOKA? That was our next test. Actually, it does. When we asked ChatGPT about HOKA directly, it knew the shoes well. The product knowledge is definitely there.
When the choice was restricted to Nike versus HOKA, the result was much closer. Nike received four clear buying preferences, HOKA received three, and the remaining three answers depended on the runner. So when HOKA is named, ChatGPT sees some of the same comfort strengths that real customers talk about. But when the field is open, HOKA often doesn’t come up at all.
So where is ChatGPT getting the information that shapes those shortlists? In ten questions about general running-shoe evidence, Nike’s own website was cited in three. HOKA’s website wasn’t cited at all. Independent sites and research databases like RTINGS were cited in eight responses. AI can get evidence from the brand, from reviewers or from independent research. If a brand leaves information out, other people end up telling its story.
Let’s pull all the clues together. Nike is vastly bigger but growing more slowly. HOKA is smaller, growing much faster and has passionate advocates. Both have legitimate marathon racers. Yet ChatGPT introduces Nike far more often.
Guys, what advice can we give HOKA? What would you actually change?
First, make the facts consistent. Use one brand name everywhere: HOKA, not the old name HOKA ONE ONE, which just dilutes the results. Label clearly which shoes are for racing and which are for training. Fill in the missing specifications, and make sure people in each country land on the right website.
Second, make genuine customer evidence easier to read, including independent reviews and real customer feedback, presented accurately.
Third, give independent testers the information they need, like stack height, so they can compare HOKA’s real strengths fairly.
Exactly. HOKA’s strongest case may be more specific: certain runners looking for cushioning and a smooth ride. Test those buying situations and make sure the evidence supports the claim. Don’t try to win every question. Compare the results with the original baseline and refine the approach using evidence. An improvement after a change isn’t automatically proof that the change caused it; that depends on the test design. The goal is a repeatable process for measuring discovery, understanding, trust and preference.
HOKA shouldn’t just tell every AI system that it has the best marathon running shoe. The key is to bring the whole ecosystem of information and evidence together, to help AI tools like ChatGPT find, understand, trust and choose them.
That’s what makes this such an interesting case study. If HOKA doesn’t come up in the initial search, customers might go straight to a competitor’s store and buy the shoe ChatGPT says is best. If they do end up in a store choosing between Nike and HOKA, the salesperson could recommend HOKA. But if the customer asks ChatGPT for its opinion, it may talk them out of it. And we all know people are going to use AI tools like ChatGPT to help them make these decisions more and more often.
Yes. The difference between being known, being understood and being recommended by AI tools is becoming business-critical, for running shoes and for almost any business.
So, did you guess right? Let us know in the comments. Next, we’re off to Seattle to see which hamburger AI tools like Google AI recommend. See you then.