Nike launches its first mini app inside ChatGPT with OpenAI: the running shoe finder

When you ask ChatGPT how to train for your first 10K and now, in that same conversation, you can ask it to recommend running shoes and buy them without leaving the chat, something has fundamentally shifted in how running footwear is sold (and chosen). That's exactly what Nike just launched alongside OpenAI: the Running Shoe Finder, their first mini app within ChatGPT, available for now only in the United States.

The Context

"Mini apps" within ChatGPT are a relatively new OpenAI feature: they allow brands and developers to build interactive experiences that live inside the conversation, without needing to open another tab, another app, or the brand's website. Instead of navigating through categories, filters, and reviews on Nike.com, the user simply converses: "I run 5K three times a week, I have mild overpronation and I'm looking for something for tempo runs" — and the assistant returns targeted recommendations, with context, comparisons, and (eventually) the ability to purchase right there.

For Nike, this is a big bet. The brand has been rebuilding its relationship with the direct consumer (DTC) after years of ground lost to competitors like Hoka and On, and jumping into the ChatGPT ecosystem—which already has hundreds of millions of weekly users—is a way to be present at the exact moment someone is deciding what to buy, without the friction of their own websites or apps. It's, at its core, the same move brands made 15 years ago when they started selling on Amazon: go where people already are, not where you'd like them to be.

The real question isn't whether the technology is novel (it is), but whether conversational AI can replace what historically required a specialist at a running store: understanding your gait, your volume, your racing goal, and your injury history to suggest the right model.

What the Science Says

Here it's worth being honest about expectations. Choosing shoes isn't a problem that gets solved by text conversation alone, no matter how smart the algorithm is. Evidence in biomechanics and injury prevention shows that individual factors—gait type, bone injury history, training volume, surface—interact in complex ways to determine what footwear reduces (or increases) injury risk. The review by Warden and colleagues on management and prevention of bone stress injuries in distance runners (2014) is clear that there is no universal "perfect shoe": shoe prescription should consider cumulative load, training surface, and individual history—something a chatbot can approximate with good questions, but doesn't replace a real biomechanical assessment.

Similarly, the work by George and colleagues (2024) on return-to-running criteria following bone stress injuries emphasizes that footwear and training load decisions in runners with an injury history require specific clinical criteria, not just preference or subjective comfort feedback. In other words: AI can be an excellent starting point for someone healthy who is beginning or wants to refresh their rotation, but it doesn't substitute for a sports podiatrist or physical therapist when pain, previous injury, or atypical biomechanics are involved.

⚖️ PMP Verdict

It works for the recreational runner who wants quick initial guidance, without sales pressure, to refresh shoes or try something new within a reasonable range of Nike options. It's useful, fast, and probably better than guessing just by looking at photos on a website.

It doesn't work for (yet) someone with an injury history, complex biomechanics, or about to make a significant volume jump (marathon, ultra). There, no conversation with AI replaces an in-person gait analysis or the opinion of a professional who sees you run.

What It Means for You

If you try the Running Shoe Finder, get the most out of it by providing real context: your weekly volume, usual surface, racing goal, and any past discomfort or injury. The more accurate information you give it, the better the recommendation. But treat it as an initial filter, not a final verdict: try on the shoe, run with it progressively (never debut it in a long race), and pay attention to your body's signals in the first few weeks. Technology can shorten the search; the final decision is always validated by your own training.

References

• Warden SJ, Davis IS, Fredericson M. "Management and prevention of bone stress injuries in long-distance runners." J Orthop Sports Phys Ther, 2014. https://pubmed.ncbi.nlm.nih.gov/25103133/
• George ERM, Sheerin KR, Reid D. "Criteria and Guidelines for Returning to Running Following a Tibial Bone Stress Injury: A Scoping Review." Sports Med, 2024. https://pubmed.ncbi.nlm.nih.gov/39141251/