Brands now direct roughly 39 percent of their total advertising spend through retail media networks, according to Bain research. That is an extraordinary concentration. Roughly two dollars in every five that a consumer goods company spends on advertising flows to a channel whose core promise, that it can prove a sale happened because of the ad, rests on measurement infrastructure that has not kept pace with the spend.
The position here is simple: retail media is over-funded relative to its demonstrable, incremental return. The asset class has benefited from a story that sounds airtight but is not. And the current macro environment makes the mispricing dangerous.
The Closed-Loop Illusion
Retail media networks sell access to first-party purchase data. The pitch is that, unlike a billboard or a social post, a sponsored listing at a major grocer can tie ad exposure directly to a basket. That sounds like the attribution problem is solved.
It is not. Closing the loop between an ad impression and a purchase is not the same as proving the purchase was caused by the ad. A shopper searching for oat biscuits on a grocery app who sees a sponsored result for oat biscuits and then buys them was almost certainly going to buy oat biscuits. The network records a conversion. Your incrementality number should be close to zero. In many retail media audits, brands discover exactly this: the channel is harvesting demand they already owned.
The measurement gap is not a niche concern. Incrementality testing, the practice of holding out a matched control group from ad exposure and comparing purchase rates, is still not standard practice across most retail media buys. Rate cards are set, budgets are approved, and campaigns run before anyone has established a baseline for what "working" actually means.
A Concentrated Spend in a Deteriorating Consumer Environment
The macro context makes this mispricing more urgent, not less. Consumers across major markets are under real financial pressure. Kraft Heinz CEO Steve Cahillane has said that shoppers are "literally running out of money at the end of the month," with some households dipping into savings, as reported by ConfectioneryNews. PepsiCo CFO Steve Schmitt has warned that the Iran conflict is likely to fuel another round of inflationary pressure across the consumer economy, per the same reporting.
At the same time, Alvarez and Marsal survey data shows that 42 percent of U.S. grocery shoppers plan to switch to less expensive stores this spring, up from 31 percent last fall, according to Food Dive. Over half of those moving to discount stores say they plan to buy the same brands they always have. So brand loyalty is not the first thing to break. But footprint is shifting. And if a meaningful share of your target shopper is moving to a discounter where your retail media network has limited reach or does not exist, your sponsored listings at a premium grocer are reaching a shrinking audience.
Indian FMCG brands provide a useful parallel under pressure. Dabur India faces 10 percent inflation this fiscal year and has implemented a 4 percent price increase, as reported by The Economic Times. Britannia faces nearly a 20 percent rise in fuel and packaging costs and is preparing selective price increases and pack-size reductions, per the same reporting. HUL has seen 8 to 10 percent cost inflation on materials, according to the same source. Brands facing that kind of cost pressure cannot afford to park 39 percent of their ad budget in a channel they cannot measure properly.
The Retailer's Incentive Is Not Your Incentive
Here is something worth sitting with. Retail media networks are, in most cases, owned and operated by the retailers whose shelves you also pay to be on. The retailer has a strong incentive to show you that the ad drove the sale. Their attribution model, their data, their reporting interface. The conflict of interest is not hidden, but it is routinely ignored at budget time.
This does not mean the channel has no value. It means the value is very hard to see clearly from inside the system. The retailer wants high CPMs and expanding network budgets. You want incremental sales at an acceptable cost. Those two things overlap sometimes and diverge a lot.
The structured equivalent outside retail media is trade promotion. Brands spent decades over-investing in trade promotions before rigorous post-event analysis showed that a large share of promoted volume was simply forward-bought by shoppers who would have purchased anyway, or cannibalised from adjacent SKUs. Retail media is repeating that cycle at speed, with better branding and worse transparency.
The Objection: But My Retail Media ROAS Looks Good
The strongest counter-argument is the one you will hear in every internal review: the return on ad spend figures from retail media campaigns look strong. Some do. Sponsored search at the point of a confirmed purchase intent can genuinely lift consideration among undecided shoppers. For a new product launch, where there is no pre-existing demand to harvest, the incrementality picture is more plausible. Display and video formats higher up the retail media funnel have a much weaker incrementality case and often carry CPMs that compare poorly to alternatives.
The problem is not that retail media never works. The problem is that the blended ROAS figure your network partner reports is almost certainly inflated by demand-harvesting that would have converted with no ad spend at all. When you cannot separate the two, you are paying a premium for claimed performance that includes a large amount of activity you already owned.
The discipline required is incrementality holdout testing, run consistently, reported honestly, and used to set budget. Most brands are not doing this. The ones who are tend to find that effective retail media spend is a fraction of their current commitment.
What You Should Do Next Week
First, ask your retail media partners for their incrementality methodology before your next budget review. Not their ROAS figure. Not their attribution model. Their holdout testing protocol. If they cannot produce one, that tells you something important about how your money is being measured.
Second, run a holdout test on your largest current retail media line item. Pick a geography or a shopper segment where you can suppress the ad without losing shelf space, measure purchase rates against the exposed group, and calculate the true incremental lift. The number will almost certainly be lower than what your dashboard shows. Use it to right-size the allocation.
Third, redirect a portion of the freed budget toward activity you can test against a genuine alternative, branded search, targeted digital video, or in-store activation at the discounters where your shoppers are increasingly spending. The Alvarez and Marsal data showing 42 percent of shoppers planning to move to cheaper stores is a signal that reach outside premium grocer networks matters more now than it did a year ago.
Fourth, when you present the reallocation internally, frame it accurately. Retail media is not broken. It is overpriced relative to what it demonstrably delivers, in a cycle where every dollar of ad spend needs to earn its place. That is a defensible commercial argument, and it is one you can back with data from your own tests rather than from the network's reporting interface.
The brands that will look smart in 18 months are not the ones who pulled out of retail media entirely. They are the ones who stopped treating a retailer's attribution model as an independent audit and started demanding proof before paying the rate card.