MadTech Daily: Digest: Brits Turn to Commercial Media Alone; ISBA Study Examines UK Retail Media Networks

Retail Media's Measurement Problem Isn't Getting Fixed

A new ISBA (Incorporated Society of British Advertisers) study confirms what anyone running a brand media budget already knows: UK retail media networks handle their own first-party data competently, but their metrics don't line up across networks. A "sale" at Tesco doesn't mean the same thing as a "sale" at Boots. Attribution windows vary. Conversion definitions vary. You can't compare them.

This is not a UK problem. Digiday's reporting on 11 US retail media networks pitching advertisers in New York finds the same thing.

Why it stays broken

The networks that could fix this are the ones who profit from it staying broken. Each network grades its own return on ad spend, books the credit, and reports numbers nobody else can check. A common conversion definition would let brands rank networks head to head. Ranking creates losers. The weaker performers have every reason to keep the fog, and they have the leverage to do it.

The ISBA finding that networks "manage their own data well" actually gives them cover: it tells them they're already competent, so the only remaining item to standardize is the part that would expose them to comparison.

What actually happens on the buy side

A brand analyst trying to build one dashboard across five retail media networks hits a wall fast. One network counts a view-through at 30 days, another at 7. One calls a click a conversion, another needs a completed basket. You can't sum them, so you don't. The analyst picks the two networks with the cleanest export and routes around the rest in the mix model. Fragmentation doesn't get escalated. It gets absorbed. Networks that make comparison hard don't lose the budget. They just stop being measured, which for a weak performer is the better outcome.

Brands and agencies have already lowered their expectations to match. They've stopped asking for full comparability. They want enough consistency to defend a budget split to a CFO, which is a lower bar than the industry keeps promising to clear.

Our call: Amazon Ads, Walmart Connect, and Kroger Precision Marketing will not voluntarily adopt a common conversion definition and shared attribution window before the 2027 IAB retail media guidance cycle closes. The incentive to stay uncomparable is strong, no external deadline forces the issue, and the largest networks have had the leverage to deliver a standard for years and haven't used it. The only path that actually moves this is a top-10 advertiser writing common definitions into its contracts and dragging networks along one renewal at a time. Slower than a standard. Probably the only thing that works.

If you want a fast read on where any given network stands: ask for its exact conversion definition, its attribution window, and whether it will accept a third-party measurement partner reading the same event. The ones that say yes would survive the comparison. The ones that stall are answering the question.

Your draft

A 97-second AI-hosted digest gave us three items. Two are geography-locked news blips. One is a real problem that keeps showing up no matter how many times the industry promises to fix it: retail media networks measure themselves however they like, and brands can't compare them.

That's the decision worth framing. The ISBA (Incorporated Society of British Advertisers) study says UK retail media networks handle their own first-party data fine, but the metrics don't line up across networks, so a brand can't tell whether Tesco's version of a "sale" means the same thing as Boots'. This is not a UK problem. Digiday's own reporting on 11 retail media networks pitching advertisers in New York says the same thing in a different accent.

How hard is this to undo? For an operator, easy. Nobody has to bet the business today. This is a "where do I put standardization pressure and budget" call, and you can change your mind next quarter.

What's actually being decided: whether you keep treating each RMN as its own walled measurement garden, or push (through your agency, through the IAB, through your own contracts) for common definitions before you commit more spend.

What sets the deadline: nothing external. Which is exactly why it never gets solved.

---

The Market Analyst. Retail media is the fastest-growing money in the industry, and it's growing precisely because it can't be compared. Each network grades its own homework, reports a return on ad spend nobody else can check, and books the credit. Standardized metrics would let brands rank networks head to head, and ranking creates losers. The networks with weaker real performance have every reason to keep the fog. For an outsider: the people who could fix the measurement mess are the same people who profit from it staying messy. That's why an ISBA study lands, gets nods, and changes nothing.

The Skeptic. Standardization has been "urgent" in retail media for three years running. Every trade body issues a framework. Adoption stalls. The ISBA finding that networks "manage their own data well" is doing quiet damage: it tells the networks they're already competent, so the only thing left to standardize is the part that would expose them to comparison. Why would Walmart Connect or any large network volunteer into a system that lets a brand say "Kroger beat you on incremental sales"? They wouldn't. The blocker isn't capability. It's that transparency is a competitive giveaway for whoever's behind.

The Operator. Tuesday morning, a brand analyst tries to build one dashboard across five RMNs. Network A counts a view-through at 30 days, Network B at 7. One calls a click a conversion, another needs a basket. You can't sum them, so you don't. What actually happens: the analyst picks the two networks with the cleanest export and quietly ignores the rest in the mix model. Fragmentation doesn't get escalated. It gets routed around. The networks that make comparison hard don't lose the budget, they just stop being measured, which for a weak performer is the better outcome.

The Customer / End User. The brand and agency are the customers here, and they've stopped asking for perfect comparability because they've learned it isn't coming. What they actually want is enough consistency to defend a budget split to a CFO. That's a lower bar than the industry keeps pitching. They don't need one currency across all of retail media. They need three or four definitions that hold still long enough to model. On the solo-consumption item: 44% of UK commercial media time now consumed alone, up hard among young adults per the IPA. Co-viewing math on connected TV gets shakier. If your CTV pitch still leans on household reach multipliers, the audience on the other end is increasingly one person with headphones.

The CFO. Money follows what you can measure, and measurement fragmentation quietly caps how much a rational finance team will move into retail media. If I can't compare networks, I can't optimize across them, so I anchor to the one or two I trust and starve the rest. The networks reading this as "we're winning by staying opaque" have the incentive backwards at the category level: the fog isn't holding total retail media spend down much yet because the channel is close to the purchase. But the day a common ROAS definition arrives, budget reallocates fast toward whoever actually performs. The strong networks should want standardization. The fact that they're not pushing hard tells you who thinks they'd lose.

---

Where the council splits. The CFO thinks standardization is coming and will reward the strong networks, so the smart ones should lead it. The Market Analyst and Skeptic think it isn't coming precisely because the people with the power to deliver it are protecting their grades. That's the real disagreement: is opacity a temporary friction or the business model?

Second split: the Operator says fragmentation gets quietly routed around, so nothing forces a reckoning. The Customer says brands have already lowered their demands to match. Both point the same direction. Nobody's angry enough to break the standoff.

What it hinges on. One belief: do the largest retail media networks gain or lose from letting brands compare them directly? If the biggest, best-performing networks decided common metrics would win them share, we'd have a standard by now. We don't. Which means the networks with the leverage don't believe they'd win the comparison, or they'd rather keep the credit they book today. The council leans hard toward "opacity is the model, not a bug in it."

What to verify before you lean on any RMN framework: ask each network for its exact definition of a conversion, its attribution window, and whether it will accept a third-party measurement partner reading the same event. The ones that say yes are telling you they'd survive the comparison. The ones that stall are telling you the answer to the whole question.

---

Prediction: No cross-network retail media measurement standard with a common conversion definition and attribution window will be voluntarily adopted by the three largest US retail media networks (Amazon Ads, Walmart Connect, Kroger Precision Marketing) by the time the 2027 IAB retail media guidance cycle closes, despite continued trade-body pressure from the IAB and bodies like ISBA.

Confidence: Medium. The incentive to stay uncomparable is strong, but a large brand could force it through contracts.

Why: Retail media grew fastest while each network graded its own return on ad spend, and a common standard would let brands rank networks head to head, which creates losers among the weaker performers. The ISBA study praising networks for managing their own data "well" hands them cover to resist standardizing the one thing that matters, comparability. The largest networks have the leverage to deliver a standard and haven't, which is the clearest evidence they don't think they'd win the head-to-head or would rather keep booking their own credit. The opposite outcome, voluntary adoption, would require the strongest networks to invite direct comparison against rivals with nothing forcing them to, and no external deadline exists to make them.

Revisit by 2027-06-30: We're right if Amazon, Walmart Connect, and Kroger are still reporting on their own metrics with no shared conversion definition or accepted third-party read across all three. We're wrong if two or more adopt a common measurement framework, or accept a single third-party measurement partner applying identical definitions across their networks, within that window.

The one crack in this: a top-10 advertiser big enough to matter could write common definitions into its contracts and drag the networks along one renewal at a time. That's slower than a standard, and it's the only path I'd actually bet on.