I recently ran into a boxing problem. I have an existing wrist issue, so I needed longer hand wraps for additional wrist support.
I started on Amazon, but the longest wraps I found still weren’t long enough.
Instead of asking ChatGPT for a product recommendation, I simply asked whether longer boxing wraps existed.
It recommended Rival Mexican Hand Wraps.
I wanted to know why.
Investigating the Recommendation
I looked at ChatGPT’s cited sources.
The primary citation was Title Boxing, specifically a product page for extra-long hand wraps.
Interestingly, although Title Boxing was the primary cited source, ChatGPT didn’t recommend buying from Title Boxing. Instead, it recommended a Rival product and linked me to Rival’s e-commerce website.
It also cited Reddit and other boxing-related sources.
This is worth noting.
The primary cited source and the final recommendation were different.
That suggests another reasoning layer exists between the evidence ChatGPT gathers and the product it ultimately recommends.
ChatGPT’s Reasoning
I asked ChatGPT how it arrived at Rival.
It summarized its reasoning as:
- The user wants longer wraps for wrist support.
- Longer wraps solve that problem.
- Rival is a respected boxing brand known for longer wraps.
- Recommend Rival.
The recommendation wasn’t simply matching keywords.
It identified the problem, selected a solution, then chose a trusted brand.
Comparing It to Google
I then Googled:
super long boxing wraps
Rival didn’t appear in the first several organic results.
Instead, Google showed multiple brands, including several products on Amazon that met my original requirement.
Even though ChatGPT knew I had been shopping on Amazon, it still recommended Rival.
That suggests authority and corroboration outweighed marketplace availability.
A Possible Explanation
Rival already has a strong narrative within boxing.
Years of YouTube videos, Reddit discussions, product reviews, coaching recommendations, and community conversations likely reinforce Rival as a trusted brand.
In this case, the recommendation was for Mexican-style hand wraps, which carries its own narrative through boxing culture.
If AI reasons across those narratives, then it isn’t simply recommending products.
It’s recommending products supported by a broader ecosystem of corroborated narratives.
The Recommendation Didn’t Determine the Purchase
I didn’t buy the Rival product.
I went back to Amazon and bought another brand.
The reason was simple.
A boxing hand wrap is a commodity product.
My primary requirement wasn’t brand.
It was length.
Amazon already fits my purchasing behavior.
I have Prime, I buy products there regularly, and the purchasing process is familiar and frictionless.
ChatGPT influenced what I considered.
It didn’t determine what I ultimately purchased.
AI’s Justification for Product Recommendation
AI said this specifically when I asked about its recommendation.
ChatGPT: “The user wants longer wraps for wrist support. Rival is a respected boxing brand. I know they make longer wraps. Recommend those.”
One thing this implies is the brand’s credibility. There are tons of boxing wraps…
AI perceived Rival as a credible boxing brand > makes boxing wraps > found one that matched user’s decision criteria.
Establishing and maintaining credibility in a brand is not new to companies.
The challenge of having it become AI’s top of mind brand among others will be a special accomplishment.
Purchase Decision Takeaway
This case highlights two separate questions:
Why did ChatGPT recommend Rival?
and
Why did I buy something else?
Those are different problems.
AI recommendations are only one participant in the buying journey.
The final purchase is still influenced by existing habits, platform trust, convenience, pricing, and personal buying behavior.
Understanding both sides of that process may become increasingly important as AI reshapes product discovery.
Leave a Reply