The Fragmentation Problem Finally Gets a Coalition
On June 19, 2026, dunnhumby launched its network alliance, an attempt to solve the single most persistent complaint in retail media: fragmentation. Every retailer with a media network has built its own audiences, its own buying interface, and its own measurement, which means a brand running a campaign across four grocers effectively runs four separate campaigns with four sets of rules. Dunnhumby's pitch is a single route to plan, activate, and measure across multiple retailers and media owners without rebuilding the work for each one. It is a coalition play in a market that has, until now, been defined by walled gardens.
The framing came directly from the top. Josh Bottomley, dunnhumby's CEO, said brands are demanding scale, consistency, and transparency, but fragmentation has held the market back. That is the correct diagnosis. Retail media revenue has soared, yet the buying experience has remained stubbornly artisanal, with agencies stitching together incompatible dashboards and arguing over whose attribution model counts. An alliance that standardizes activation and reporting across competing retailers is, on paper, exactly the connective tissue the category has lacked. The question is whether retailers will actually cede enough to make it real.
Who Is Piloting, and in What Categories
The launch arrives with credible pilot partners rather than vague intentions. Tesco, B&Q, John Lewis, and Waitrose are taking part, spanning grocery, DIY and household, and health and beauty. That category spread is deliberate. It lets the alliance prove the model across very different purchase behaviors, from impulse grocery buys to considered home-improvement projects. Kingfisher's B&Q alone reports more than 300 stores across the UK and Ireland and over 22 million annual shoppers, while Tesco brings the scale of the country's largest grocer and one of its most developed retail media operations through Tesco Media.
The pilots begin in summer 2026, and dunnhumby has signaled that additional retailers are expected to join. We would note the strategic significance of pairing direct and adjacent competitors in the same framework. John Lewis and Waitrose sit under one partnership, Tesco competes with both on grocery, and B&Q plays in an entirely different aisle. Getting these names to test a shared standard at the same time is itself the headline. If the alliance can demonstrate that a brand can buy health and beauty across Tesco and Waitrose with consistent measurement, it will have proven something the category has only ever promised.
Kevel and the Open Model Bet
The alliance is being delivered in partnership with Kevel, the ad-tech firm known for API-first, build-your-own ad-server infrastructure rather than packaged software. That choice is telling. By leaning on a developer-oriented, composable backbone, dunnhumby is positioning the alliance as open and collaborative rather than as another monolithic platform that retailers must adopt wholesale. The stated design lets each retailer maintain control of its own media business while benefiting from shared standards for activation, measurement, and reporting. In other words, retailers keep their P&L and their data sovereignty, and the alliance supplies the interoperability layer on top.
This is the crux of whether the model survives contact with reality. Retailers have resisted shared networks precisely because retail media margins are too good to dilute and shopper data is too valuable to expose. Dunnhumby is betting that an architecture which preserves individual control while connecting purchase data to scaled publisher reach can overcome that reluctance. The use of an open ad-tech partner rather than a single dominant platform is a credibility signal to retailers wary of handing leverage to a gatekeeper. Execution will determine whether open turns out to mean genuinely neutral or simply dunnhumby-mediated.
Why This Threatens the Walled Gardens
The most interesting tension here is competitive. The dominant logic of retail media has been the walled garden, where each retailer hoards its first-party data and forces brands to come to it on its terms. The largest networks have the scale to make that work. Smaller and mid-sized retailers do not, and for them a shared alliance is a lifeline: it lets a regional grocer or specialist offer brands access to multi-retailer reach they could never assemble alone. Dunnhumby, sitting on decades of Tesco purchase data and a global analytics footprint, is uniquely positioned to broker that coalition.
For brands and agencies, the upside is obvious: less rebuilding, more comparable measurement, and a credible counterweight to the take-it-or-leave-it terms of the biggest networks. For retailers, the calculus is subtler. Joining trades some independence for reach and efficiency, and the ones most tempted will be those who cannot win the walled-garden game on their own. We expect the alliance's success to be measured less by the marquee names in the pilot and more by how many second-tier retailers it can pull in once the summer tests report results. If it scales, it reshapes the bargaining table. If it stalls, the walled gardens win by default.
