There was a window in 2023 and 2024 when nearly every national chain put out a press release about a flagship store of the future. Smart shelves, computer vision at the door, app-based wayfinding, electronic shelf labels updated by the second, RFID across categories that hadn't carried it before. By the third quarter of 2025, the post-mortems are happening — usually quietly, in operations review meetings rather than press releases — and the picture is more sober than the launches suggested.

The honest scorecard

A director of stores at a national specialty chain summarized it this way: "Out of the six things we put in the concept store, two are going national, two are staying in that one store as a marketing prop, and two we're killing." That ratio — roughly a third graduating, a third surviving as a halo, a third quietly removed — tracks closely with what we've been hearing from operators at three other chains during Q3 evaluation cycles.

The technologies clearing the bar tend to share two features: they reduce a labor-hours line item that store managers were already complaining about, and they don't require the customer to learn a new behavior. The ones being killed tend to require either expensive ongoing IT support or a level of customer adoption that the pilots never produced.

What's earning a rollout

  • Electronic shelf labels. ESLs were the most-piloted technology of 2024 and are the most-rolled-out in 2025. The business case isn't dynamic pricing — most chains are still cautious about visible price changes during the shopping trip — it's price-accuracy compliance and the labor cost of paper tag changes. A category manager at a grocery chain told us their ESL rollout paid back primarily through the elimination of tag-printing and tag-hanging hours, not from any pricing strategy.
  • RFID in apparel and footwear. Inventory accuracy gains in the 90%-plus range, against a starting point in the 60s and 70s for many chains, have made the case decisively. The rollout question now is who pays — the brand or the retailer — and that's still being negotiated case by case.
  • Computer vision at self-checkout. Not the full Just Walk Out vision, which has retreated, but the narrower use case of catching mis-scans and ticket-switching at self-checkout lanes. Several grocers and mass merchants are scaling these in 2025 after pilots showed shrink reductions in the meaningful single digits.

What's stalling

The list of technologies that are not graduating to national rollout is longer than the list that is.

Smart shelves with weight sensors have largely been pulled from pilots. The maintenance burden — recalibration, false alerts, sensor failures — outran the operational benefit, and operators we spoke to said store associates lost faith in the data within the first quarter.

In-store augmented reality, whether through customer phones or fixed displays, continues to demo well and convert poorly. A digital director at a beauty chain put it directly: "The AR mirror gets used by influencers and almost nobody else."

Robots — the inventory-scanning kind that rolled the aisles overnight — are mostly gone or being replaced by fixed cameras and handheld devices used by staff. The robots could see the gaps but couldn't fill them, and the data they produced often duplicated what existing systems were already showing.

The pattern underneath

What ties the survivors together isn't sophistication. It's that they replace a clearly costed operational task rather than try to create a new customer experience. The pilots most likely to be killed in 2026 review cycles, based on what operators are telling us, are the ones still being justified on the basis of "delighting the customer" or "future-proofing the store" without a labor line attached.

For 2026 planning, the practical takeaway from Q3 evaluations is that the bar for store technology has moved. A pilot that worked in 2023 — meaning it shipped and customers used it — would not necessarily survive a 2025 review if it didn't change a labor or shrink line. The chains that have been most disciplined about this are now ahead, not because they piloted less, but because they killed faster.