Thane_Bitter I feel like this (not just online search but retail in general) is something that could REALLY benefit from hiring and working with teams of AI experts. General merchandisers and inventory folks are great at making sure stores don't run out of X, and buy Y that is being pushed by their suppliers, but when you have a store that carries 6000 SKU's, having a great insight into what's the next big thing and what's wrong and what doesn't match your target markets and customers, sure you can keep your finger on the high movers and have insight, but damn... AI is designed to look at billions of data points and make reasonable decisions and fix things that seem "wrong". A reasonable approach should be "take the last 100,000 or million searches on our website for products, review the recommendations shown to our customers, review what they actually purchased, cross reference that to every product sku and what searches led to the most sales of those products, and recommend what proper categories each item should be in, what the trend in sales of each product is, what items commonly pair with each other, what items are commonly purchased as one offs, what items are most often purchased by first time visitors and return visitors, and make reasonable recommendations on inventory, presentation, marketing, and market segmentation for everything we carry". (Yes it would be a lot more work than a llm search based on that, that's why you need a team of AI developers in-house in your IT/IS team to build tools that do exactly that - and have those tools provide daily recommendations to the folks who are implementing it).
