What’s Holding Back AI Adoption in Retail, Beyond the Technology

Retail isn’t short on technology today. There’s no shortage of vendors, success stories, or retailers adopting new tools: testing them, running pilots, evaluating providers. The gap shows up a step later, between wanting to try a technology, or even piloting it, and getting that same technology to actual, network-wide use.

Part of that gap comes from the pressure to show progress fast. At the executive level, any AI investment also works as a signal to the outside world, not just an internal bet. That pressure to show innovation can cloud the what and the how. That’s exactly why it pays to look beyond the technology itself: the conversation is moving away from the dashboard, analytics, and strategy toward something more practical. That shift already shows up on the ground: operations teams are asking for technology that translates into something they can apply on the sales floor, in the actual execution of day-to-day work.

What explains that gap in more detail? A recent survey by Virtuous AI and Chief Executive Group, covering more than 300 CEOs at mid-market companies in the United States, shows exactly where the bottleneck sits. 98.5% of those executives say AI has already generated value for their business. Only 7% have a company-wide AI strategy with multiple initiatives underway. 52% are still in the pilot phase, and 31% explored AI without ever implementing it. Put those three numbers together, and the message is clear: conviction about the value is settled, the ability to carry that value across the whole operation is not. The CEOs themselves name the obstacles, and not one of them is a software problem: 86% point to a lack of in-house expertise, 81% to difficulty integrating AI with existing systems, and 65% to data quality and access problems.

The same picture shows up on the retail side. A monday.com survey of more than 1,800 retail executives found that 96% are already implementing, piloting, or exploring AI agents. Practically no one is sitting out the conversation. At the same time, 61% question the consistency and quality of AI’s output, 45% cite integration complexity, and 35% cite resistance from their own teams. All three are operational and people-related barriers, not technology ones.

AI pilots rarely cross the line between the office and the store. One born in a corporate committee tends to stay there: it gets measured against metrics from that same level, and reviewed by the same team that proposed it. For a pilot to actually be worth anything, its learnings need to reach the hands of the people working the floor.

Faced with that problem, some chains are taking a different path. Instead of adding another layer of approval before making a decision, they’re removing one. Coresight Research documented that several retailers are flattening entire management layers, aiming to move AI decisions closer to the store team and, even closer, to associates and customer-facing roles. The logic is simple: every approval level a decision must clear is one more step away from actual execution and one more day before it sees real use on the floor.

Javier, Frogmi’s Head of Sales, sees it come up in conversation after conversation with retail clients and prospects: in almost every one, the same question surfaces, how to actually get AI to reach the store instead of staying stuck at headquarters.

“There’s no shortage of AI tools out there today. That stopped being the problem a while ago. The real challenge is getting them onto the sales floor as something useful and simple, something the team uses without a second thought,” Javier sums up. “The teams that actually get value out of it aren’t the ones that buy the most sophisticated tool. They’re the ones that turn it into something their people use every day.”

The same pattern repeats across all three sources: what’s holding back AI adoption in retail today is that the barrier shifted from technology to organization.

Putting that into practice means thinking concretely about which tools frontline staff actually get to support their daily work, not just installing tools that promise to deliver value. Before adding anything new, it’s worth looking at which tasks eat up the most time for the store team today, and how to solve them more efficiently. And, of course, confirming which of those tasks actually move the needle for the business.

Closing that gap in decision-making means combining the strategic view of whoever makes decisions with the hands-on experience of what actually happens day-to-day on the floor. Put those two perspectives together, and you get a complete decision. AI is already in use. Now it has to actually add up on the sales floor. We’ve already written about how to make retail technology adoption change your stores: the same principle applies here.