🔥Join eCom Hot Sauce newsletter if you want to get such hot news and tips before the other eCom peeps!
Amazon sellers have spent 20 years fighting for the Buy Box.
Now AI commerce may be creating another one.
ChatGPT’s newer shopping experience can highlight a single retailer alongside a product and price, effectively creating something that looks a lot like an AI-powered Buy Box.
And according to ecommerce researcher Juozas KaziukÄ—nas, there’s an important catch:
Retailers supplying structured product feeds appear to have a major advantage.
That could have huge implications for ecommerce brands.
Because the next battle might not be:
“How do I rank #1 on Google?”
Or:
“How do I win Amazon’s Buy Box?”
It might be:
“How do I become the retailer ChatGPT puts on that button?”
AI Doesn’t Necessarily Pick the Cheapest Offer
KaziukÄ—nas tested several products and found examples where ChatGPT’s highlighted retailer wasn’t the retailer offering the lowest visible price.
In one Moccamaster example, the default shopping option pointed users toward the manufacturer’s website even though Bloomingdale’s appeared to offer the product for $10 less, and Amazon reportedly had the lower price too.
Another example involving a Breville coffee maker highlighted an even bigger issue:
Product matching.
A lower-priced offer appeared during comparison, but closer inspection suggested it was actually a different product variant.
That shows how messy AI commerce can become.
Price means nothing if the system is comparing two slightly different SKUs.
Product Feeds Are Becoming Digital Shelf Space
For years product feeds felt like boring ecommerce plumbing.
Something the technical person dealt with.
Google Merchant Center.
Facebook catalogs.
Marketplace feeds.
SKU attributes.
Variant IDs.
Yawn.
AI shopping changes that.
If an AI system uses structured merchant feeds to decide which products and retailers appear in shopping experiences, feed quality suddenly becomes a revenue issue.
Bad product data doesn’t just make your catalog ugly.
It could make your product invisible.
Missing attributes could hurt matching.
Incorrect variants could send customers to the wrong SKU.
Outdated pricing could destroy trust.
Incomplete availability data could mean the AI simply chooses somebody else.
Product feeds might become the new shelf space.
Being Recommended Isn’t Enough
There’s another important distinction emerging.
Brands are currently obsessed with AEO and GEO:
“How do I get ChatGPT to mention my brand?”
Great question.
But getting mentioned isn’t the finish line.
Imagine ChatGPT says your coffee machine is one of the best options.
Victory?
Not necessarily.
The shopper then sees a product card.
Then a retailer.
Then a price.
Then a purchase button.
Every step is another mini competition.
Winning the AI recommendation but losing the transaction link is basically the ecommerce equivalent of ranking #1 on Google and sending the customer to your competitor’s checkout.
Not ideal.
Welcome to Feed Optimization
So here’s the boring homework ecommerce brands should start doing now:
- Clean your product feeds.
- Standardize product titles.
- Make variant information crystal clear.
- Keep prices current.
- Keep availability current.
- Use consistent identifiers.
- Provide detailed attributes.
Make sure machines can easily understand the difference between model A and model B.
That might not sound as exciting as “10 Secret ChatGPT Prompts That Will Change Your Life.”
But it could make you considerably more money.
AI shopping is still evolving, and today’s interfaces and ranking logic will almost certainly change.
But the direction is pretty obvious.
AI assistants are moving closer to the transaction.
And once AI decides which retailer gets the click, ecommerce brands will have another algorithm to feed.
Literally.
🔥Join eCom Hot Sauce newsletter if you want to get such hot news and tips before the other eCom peeps!
Â