Almost every week, a major brand publishes a glowing case study about how generative AI transformed their operations. That is why Walmart’s recent admission is so refreshing. They tested ChatGPT to handle customer checkouts, and the results were brutal. The conversational AI checkout converted three times worse than their standard website.
We need more of this honesty right now. Everyone is still figuring out how to build with these tools, and public failures are just as instructive as the wins. What Walmart experienced here is a classic collision between technological novelty and deeply ingrained consumer behavior.
For the last two decades, the e-commerce industry has spent billions optimizing the traditional shopping cart. We know exactly where the buttons should go, how to minimize clicks, and how to visually signal security at the exact moment of purchase. Shoving a conversational interface into the bottom of that carefully engineered funnel introduces friction. Talking to a bot to finalize a transaction might sound futuristic in a pitch deck, but in practice, it is slower and far less intuitive than simply clicking a familiar “Buy Now” button.
Consumers trust established pathways. When they are ready to hand over their money, they do not want to chat. They want speed, clarity, and the predictable flow of a standard checkout screen. AI introduces variability into a process where the user specifically craves certainty.
Stop forcing conversational AI into your transaction layer just to check an innovation box. If you want to deploy these models effectively, point them at the top of the funnel. Use them to help customers discover products, compare complex features, or answer highly specific research questions. But when the customer finally decides what they want to buy, get the AI out of the way and let your traditional, highly optimized checkout do what it was built to do.
Source: Walmart says ChatGPT checkout converted 3x worse than its own website
