Kroger told analysts and investors on 20 October 2026 that it is running RFID in roughly 1,500 of its stores and intends to have it in more than 2,300 by 2029. The US grocery chain operates around 2,700 supermarkets across 35 states and describes itself as about twice the scale of the next largest US supermarket operator, so the 2029 target would cover the large majority of an estate that is itself due to grow by around 110 new stores over the same period.
The numbers sit in the connected store section of Kroger’s 2026 investor update presentation, a 92 page deck dated 20 October 2026. The store operations material was presented by Mark Ibbotson, executive vice president and chief store operations officer. The deck groups RFID with three other in-store technologies: robotics, at around 900 stores as of October 2026 and 2,100 or more by 2029; digital shelf labels, which are electronic price tickets on the shelf edge that can be repriced centrally instead of being printed and clipped on by hand, at around 1,100 today and 2,400 or more by 2029; and an AI and retail insights layer sitting on top. The slide lists the value creation drivers as “Labor Productivity”, “Shrink Improvement” and “Retail Media Monetization”, under the strapline “Store systems give the signal. AI prioritizes what matters. Associates get a clear action.”
Bakery is carrying the business case
One department gets a slide of its own, and it is bakery, described in the deck as “now across 1,500+ stores”. The results claimed “in one department” are:
- around 70 per cent faster inventory counts
- earlier expiration detection, leading to around 50 per cent lower waste costs
- around 200 basis points of sales lift and around 50 basis points of category share gain
A basis point is one hundredth of a percentage point, so that is a sales increase of roughly 2 per cent and a share gain of roughly half a percentage point in the bakery category.
The deck itself does not credit those figures to RFID specifically. It presents them as the connected store “working together”, which is to say tags, shelf labels, robotics and AI as a bundle. The RFID link comes from the spoken presentation rather than the slides: reporting on the 20 October 2026 remarks, Supply Chain Dive wrote on 27 October 2026 that Kroger presented its bakery RFID results and added a figure the deck does not carry, a 98 per cent read performance, meaning the share of tagged items the system successfully reads and identifies during a count. Ibbotson’s own summary, as quoted in that account: “That is a better store routine, less counting, better information, lower waste, higher availability and of course, more sales.”
One wording point worth flagging for anyone comparing coverage. Supply Chain Dive describes the bakery work as a pilot. At roughly 1,500 of around 2,700 stores it has passed that stage: more than half an estate is a deployment with a next phase attached, not a trial.
Waste is the number Wall Street will hold them to
Waste gets its own slide, headed “a key benefit to the connected store, supercharged by AI”, with a three step logic of connect the data, identify the risk, direct action, and a target of around $500m in savings by 2029. That figure sits inside a broader $4bn structural cost takeout Kroger says is enabled by technology and AI, with capital expenditure rising from about $3.9bn to around $5bn. Chief executive Gregory Foran framed the lot in one line: “Every single one of these actions has the same goal: take cost and complexity out, put the savings back into customer value and execution, and expand margins over time.”
It is worth noting what did not happen. RFID made the investor deck, but Kroger’s own press release on the day covered financial targets only and did not mention it. Chief financial officer David Kennerley’s supply chain material covered a new automated site in Franklin, Kentucky, serving the Nashville and Louisville divisions and due for completion in 2028, with no tagging detail attached.
Fresh food is the hard case for item-level tagging
Most retail RFID to date has been apparel and general merchandise, where a unit has a stable identity, a long shelf life and no expiry date. Bakery is the opposite: short dated, high waste, repriced through the day, and produced in volumes that change by the hour. That is precisely why item-level tagging pays differently there. A tag that carries a unique serial lets a store tie a specific loaf to a specific bake time and a specific sell-by date, which turns markdowns and withdrawals from a clipboard exercise into a query. The roughly 50 per cent reduction in waste costs and the earlier expiration detection are the same mechanism described twice.
On the evidence presented, this is the largest fresh food RFID deployment in US grocery that we are aware of, and the plan is to widen it. Ibbotson: “The next job is to use trusted item level data to improve replenishment, production markdowns and fresh decisions across more of our fresh [category].” Kroger also says its AI tool connects 20 data pipelines and gives visibility down to store and item level, and that it is sharing computer vision data with suppliers to improve on-shelf availability.
What Kroger has not said
Neither the deck nor the Supply Chain Dive account names a frequency band, an air interface, a standard, a tag vendor or an inlay, so anyone sizing the opportunity is working blind on the hardware. Two more open questions matter operationally. First, where the tag is applied: in store at the bakery counter, or upstream at a commissary or plant, which changes the labour maths completely. Second, which fresh category comes next, since produce, meat and deli each bring different packaging and different tag economics.
A note of caution on the 2029 numbers. Kroger’s technology record is mixed. It scaled back its automated fulfilment partnership with Ocado after assessing whether the robotic sites hit performance benchmarks, cancelled a planned North Carolina fulfilment centre in 2025 at a cost of $350m paid to Ocado, and a Tennessee facility was set to close in 2026. A store technology rolled out to half the estate on the strength of departmental results is a different proposition to a greenfield automated warehouse, but a 2029 target in a 2026 deck is still a target.
Read more at https://s202.q4cdn.com/463742399/files/content_files/Kroger-2026-Investor-Update-Presentation.pdf
