Stores ring up most retail sales but earn 5% of retail media revenue

The US Census Bureau puts e-commerce at 17.1% of US retail sales for the second quarter of 2026. That leaves roughly 83% of sales happening away from websites, mostly inside physical stores. Advertising money flows the other way.

Brick-and-mortar stores account for just five percent of retail media revenues, according to an Omdia estimate that invidis reported. So the store is where most purchasing takes place, yet it is also where the smallest share of advertising gets sold.

A guide published Oct. 30 by signage integrator Fingoweb treats that gap as a plumbing problem as much as a sales one. The argument is blunt. Screens in a store sell ads when the shelf stock, the till data and the playback logs sit behind them, and none of that comes from the screen itself.

As Fingoweb defines it, in-store retail media is ad space that brands purchase within a physical store. The ads run on screens, and their results are tracked against the retailer's own data on sales, stock and footfall. Fingoweb wrote the guide for the integrators and network operators who build these networks for retail chains. That is Fingoweb's own line of business, so the guide reads like a practitioner's playbook.

The Census Bureau's sales figures and Omdia's revenue estimate aren't measured on the same basis, one covering US sales and the other retail media revenue, but they point the same way.

Why the store gap is worth closing

The economics explain why retailers keep at it. A Journal of Marketing paper on point-of-sale signage, citing earlier research, says retailers that sell their advertising space achieve profit margins of up to 90%, against an average gross profit margin of around 3%. The same paper cites a forecast that this revenue stream will reach $45.3 billion by 2030.

Scale is the other half. Walmart, the paper notes, draws 204 million weekly customer visits to its stores and delivers advertising on an estimated 170,000 digital in-store screens across the US. With margins that far apart, a dollar of in-store ad revenue would be worth many times a dollar of product sales to the retailer's bottom line. That is why a 5% share looks less like a niche and more like money left on the shelf.

Buyers who already spend want more

Among US retail media buyers who have active in-store budgets, 43% said their organization underuses that investment, according to a survey by IAB and Grocery TV. In The Shelby Report's account of the survey, those buyers named activation complexity and measurement limits as the top barriers.

That matters because these are committed buyers. A network trying to win them over doesn't need to convince them that stores influence purchases. It has to make campaigns easier to launch and easier to prove. Fingoweb's guide makes the same point: both barriers sit in software and data.

Vendors describe the same friction from the supply side. Broadsign, which says it powers more than 35 in-store retail media networks, including Woolworths, Coles, Sainsbury's and Auchan, writes that many retailers deploy screens quickly but struggle with control, content quality, governance and data visibility. That last item is the buyers' measurement complaint seen from the retailer's chair.

Three feeds behind a sellable screen

Stock, till and playback data each answer a different question. The playback log is proof of delivery: the spot ran on that screen at that time. The other two feeds tell a brand whether that delivery could have mattered.

Measurement is harder in a store than online for a structural reason. The Journal of Marketing authors point out that physical shoppers do not leave a trail of their behavior the way online shoppers do. Online attribution is already complex, they write, and establishing causal ad effects at the point of sale is harder still.

A field study in the Journal of Marketing tracked 237 advertising campaigns and 30 million shoppers and found that digital signage at the point of sale raised the likelihood of buying a featured product by 8.1%.

The study's other findings shape what the till data should be asked. The lift grew when the signage sat near the advertised product. Unlike price promotions, signage did not change how much purchasers of the product spent. Exposure also lifted sales of other products from the same brand and in the same category without pulling purchases forward, which the authors read as incremental consumption.

That halo would matter if a network reports only the featured SKU. Brand-level and category-level POS data would capture lift that a single-product report leaves on the table.

Signage software vendors now sell these feeds as integration features. Navori lists dynamic content scheduling that adjusts messaging by time, season or inventory levels, along with analytics that track engagement and conversion rates. Broadsign advertises connections to loyalty, POS and CMS systems so campaigns can be coordinated across channels.

Two meanings of inventory

The word "inventory" does double duty in this business. To a media seller, it means screen time. Broadsign pitches "real-time inventory visibility" in exactly that sense, knowing which screen is available, and lets networks sell campaigns by location, share of voice, impression or budget, including through programmatic demand.

To a store operator, inventory is what sits on the shelf. A booked slot promoting an out-of-stock product counts as delivered media and wasted money at the same time. Scheduling that responds to stock levels, the kind Navori describes, is the bridge between the two definitions, and it only works when the stock feed is live.

Faster campaign building at Ahold Delhaize USA

Activation, the other barrier buyers named, is where some large grocers are trying AI. According to Grocery Dive, Ahold Delhaize USA's retail media arm, AD Retail Media, has embraced AI as it advances its operations, with the aim of moving faster and building campaigns more precisely. The unit is overseen by Bobby Watts, senior vice president and executive lead, who spoke on a panel at the Groceryshop conference in Las Vegas on Sept. 23.

Watts told the publication his team has deployed three agents this year to test small business use cases and see whether they deliver value. One of them, Agent Klik, mines how consumers search for products and sends trending keywords and phrases to the ad operations team as recommendations. The agent suggests and people decide. Watts also described wanting to keep the human touch of the company's grocery banners as the technology spreads.

He framed the rollout as test-and-learn, with the team encouraged to play with the tools. The placement is telling for the activation problem. Agent Klik works at the front of campaign building, picking search terms, where speed pays off and a human reviewer can still catch a bad suggestion before money is spent.

The human-touch concern has a research echo, though from the shopper's side rather than the ad desk. A February 2021 Journal of Business Research study built on 25 in-depth consumer interviews identified exploitation, interaction misfit, privacy and lack of confidence as barriers to accepting technology-enabled personalization in stores. It also named humanization versus dehumanization among the paradoxes retailers have to manage.

What wiring alone won't fix

Integration isn't the whole answer. The same Journal of Marketing study found the signage effect was uneven. It was stronger for hedonic, novel and low-priced products and for popular brands, and stronger on weekends and later in the day. It also rose during favorable weather, in crowded stores, for emotional messages and when no other promotional cues were competing. A fully connected network running the wrong products at the wrong hours would measure a smaller lift, and measure it accurately.

The promotional-cue finding is the one most likely to clash with how stores already merchandise. It suggests that stacking a screen ad on an existing in-store promotion may get less out of the screen than running it alone. A network can only make that scheduling call well if it knows which promotions are live in each store.

That gives retail media leaders a sharper pitch to make. Clean data lets a network prove results, while choices about product category and daypart decide how big those results are. The test a brand will apply is concrete: for each campaign, can the retailer show that the product was in stock, the spot ran, and what sold afterward?