Retail in the age of AI vision

The future of shopping just walked into my basement.

I had no idea where the main water shutoff was in our new (but very old) house. Built in 1955, things are just… different. So, I did what any modern homeowner would do—I asked ChatGPT!

But not just with text. I turned on the video feature and asked, "Can you help me find the main water shutoff valve?"

It told me to move my phone around. Then: “Go closer to that red valve.”

I did.

"That is your main water shutoff."

Mind. Blown.

That got me thinking. If AI can do this, how will customers use it in retail?

So, I tested it again—this time, with a running shoe. I asked ChatGPT about the Nike Pegasus 36 Trail product.

Me: What can you tell me about this shoe?

ChatGPT: That’s a Nike Pegasus 36 Trail. It’s known for being a versatile trail running shoe with good cushioning and traction. The upper is breathable, and the outsole provides solid grip for trails. It’s a reliable option for mixed terrain. Are you considering it for trail running?

Me: Yeah, I am, but I’m also curious about how it holds up in cold and rainy weather.

ChatGPT: The Pegasus 36 Trail is breathable, so it’s not specifically designed for cold and rainy weather. For those conditions, a waterproof version like the Pegasus Trail Gore-Tex may be better. It offers more protection against moisture and cold. You could also consider pairing them with warm, water-resistant socks.

That’s when it hit me: ChatGPT just became my shopping assistant.

Customers will be doing this everywhere—scanning products, asking AI for insights, and making buying decisions based on what AI tells them.

The question is: Will retailers influence those answers, or will AI piece together information from scattered, third-party sources?

This is a data game

AI-powered shopping assistance is not just search and browse—it’s conversational. That changes everything.

Retailers need to think beyond traditional product pages and SEO. If they aren’t intentionally developing their data to be optimized for how AI responds to customers' natural questions, they risk having AI misinterpret their products.

Here are some suggestions for navigating this new AI-enabled consumer landscape:

• Think like AI. The internet is messy and unstructured, and AI learns from everything—reviews, forums, product pages, social media chatter. If your product data isn’t designed to flow naturally in a conversation, AI may pull from less reliable sources.
• Know how customers actually talk. Unlike a search bar, where customers type exact product names, AI-powered interactions will be messier and more casual. Someone might ask, “What’s the best waterproof running shoe?” instead of searching for “Nike Pegasus Trail Gore-Tex.” Retailers need to anticipate real-world phrasing and optimize their data accordingly.
• Follow the zeitgeist. AI reflects what the internet is saying. If a certain product is trending for an unexpected reason—or if customers describe it in a new way—retailers should adapt their product descriptions to align with how people are actually talking about it.

Retailers need to stop thinking about data like a catalog and start treating it like a script for AI-driven conversations.

What to do next

AI vision isn’t just a cool feature—it’s a fundamental shift in how customers interact with products. To stay ahead, retailers should:

• Redefine their product data for AI. Make sure AI understands products the way customers actually talk about them.
• Experiment with AI-powered training. Build AI coaching tools that help employees answer real customer questions effectively.
• Own the AI experience. Don’t just react—integrate AI into your own apps and experiences so customers come to you first.

The retailers who embrace AI now won’t just keep up. They will set the standard.

Will your organization be among them?

Kevin Ertell is the founder and CEO of Mistere Advisory.