Your Shoppers Already Use AI to Decide What to Buy. Does Your Store's Data Keep Up?
Nearly four in ten shoppers now use AI to find products, and retailers say AI isn't slowing them down; it's everything disconnected underneath it.
Core insights: A shop can add every AI feature on the market and still not know, in under five minutes, which products made money last month. Salesforce's own retailer survey says disconnected systems, not a lack of AI, are the more common productivity problem right now.
Salesforce's sixth Connected Shoppers Report, based on a survey of 8,350 shoppers and 1,700 retail decision-makers, put a number on something most store owners have already noticed anecdotally: 39% of shoppers now use AI to discover products before they buy. Among Gen Z shoppers, that number is 54% more than half.
Retailers have noticed too. The same report found 43% of retailers are already piloting autonomous AI agent tools that handle customer service, inventory monitoring, or parts of the buying journey without a human in the loop. Three-quarters of retail decision-makers said AI agents would be essential to stay competitive within the year.
That's the headline. The detail worth pausing on is buried a bit deeper in the same report: 81% of retailers say inefficient processes and disconnected technology are actively reducing productivity on their teams. Not "could be a problem eventually." Actively, today.
The gap sits behind the storefront, not in front of it
Shoppers moving toward AI-assisted discovery is a demand-side shift; it's happening whether or not a store is ready for it. What Salesforce's data shows is that most retailers' supply-side systems, the actual plumbing connecting sales, inventory, and marketing data, haven't caught up to that shift. Four out of five retailers are saying so themselves.
That gap tends to be worse, not better, for smaller independent stores than for the large retail chains surveyed. A big-box retailer with disconnected systems still usually has a team stitching together spreadsheets behind the scenes. A store running lean with a single owner or a two-person team usually has the storefront platform's built-in reports, a separate ad platform dashboard, maybe a monthly export from the payment processor, and a lot of guessing in between.
What that looks like in practice
Picture a store that adds an AI-powered product recommendation widget this quarter, because that's the visible, easy-to-explain AI feature. Meanwhile, the same store can't quickly answer which products are actually driving margin once ad spend is factored in, which regions convert but never get promoted, or which SKUs are quietly draining inventory budget month over month. The recommendation widget is genuinely useful. It's also solving a discovery problem while a bigger data problem sits untouched underneath it.
The fix isn't necessarily a bigger AI investment. It's making the data the store already has actually usable in one place sales, inventory, and ad performance pulled together instead of scattered across three logins- so ai data analytics tools can surface the pattern (this product converts well from paid search but not from social; this SKU's margin has quietly eroded over six months) instead of an owner noticing it three months too late, if at all.
Where to actually start
Before adding another AI feature to the storefront, it's worth an honest audit of the back end: can you currently answer, in under five minutes, which products made the most money last month after ad spend? Which channel brought in your most profitable customers, not just the most orders? If the honest answer is "not without pulling three reports and doing math in a spreadsheet," that's the gap the Salesforce data is describing, and it's the one worth closing first.
Shoppers have already made the shift. The stores that come out ahead won't necessarily be the ones with the flashiest AI widget on the product page; they'll be the ones whose owners can actually see, clearly and quickly, what's working underneath it.
