Retail Media CPMs in 2026: The Benchmark Table Buyers Need

The 2026 Retail Media Benchmark Table

The pattern worth noticing is not the absolute numbers, it is the spread. Amazon's sponsored search auction is the most crowded in retail media, which pushes CPCs toward the top of every category benchmark Tinuiti publishes. Instacart, by contrast, still has less advertiser density relative to its shopper base, which is exactly why performance marketers who moved budget there in 2024 and 2025 kept reporting lower blended costs.

Kroger sits at the other extreme for a different reason. Its parent data co-op, 84.51, licenses purchase-level loyalty data that lets Kroger sell precision audiences off its own site, not just impressions on its own shelf. That is closer to a data licensing fee dressed up as a media buy, and it prices accordingly. The same logic explains why Albertsons has started experimenting with scripted branded content inside retail media: once you own the shopper relationship, you can charge for attention formats a generic banner network never could.

How Retail Media Stacks Up Against Google and Meta

Put the retail media table next to a generalist benchmark and the premium is obvious immediately. LocaliQ's Google Ads benchmark report, one of the more widely cited industry-wide CPC references, has put average Google Search CPCs across verticals in roughly the $2 to $4 range for most consumer categories, while Meta's average CPMs in comparable benchmark tracking typically land between $8 and $12 depending on placement and season.

That means a Kroger onsite display buy at $18 CPM can run 50 percent above a comparable Meta feed placement, and Amazon's Sponsored Products CPC floor sits well above what most brands pay for a branded Google Search click. On a pure cost-per-impression basis, retail media is the expensive seat at the table, full stop.

The context that changes the comparison is intent, not audience. A Meta impression finds someone scrolling; a Kroger sponsored placement finds someone actively shopping the category with a payment method attached. That is also why the closed-loop sales data behind retail media matters more than the CPM line itself: the attribution models built for retail media can tie a specific impression to a specific SKU sale in a way open web display never has been able to.

It is worth noting that Google's own search revenue growth has been decelerating (the company reported growth slowing to 17 percent in its most recent quarter) while retail media keeps compounding at rates most search and social platforms have not posted in years. Buyers reallocating budget toward retail networks are not chasing a fad, they are following where growth and first-party data both point.

Is the Premium Worth It? Reading Past the CPM

CPM is the wrong lens for a channel that sells conversion proximity. The right question is not 'what does the impression cost' but 'what does the incremental sale cost, net of what would have happened anyway.' Most retail media networks will hand you a self-reported ROAS number that flatters the channel, because the attribution window and the baseline are both set by the platform selling you the ad. That is the same conflict of interest that has dogged walled garden reporting for a decade, and retail media has inherited it wholesale.

There is a legitimate case for the premium regardless. Retail media puts your ad in front of a shopper at the exact moment they are choosing between your product and a competitor's on the same digital shelf, which is a placement Google and Meta simply cannot sell you at any price. For categories with low brand loyalty and high switching, like commodity CPG or private label competition, that proximity is worth paying up for even at a $15 to $20 CPM.

The counter-case is just as real. If your product already wins on brand search and repeat purchase, stacking retail media on top of an already-loyal buyer is closer to a tax than an investment. That is the debate showing up in how 30 marketing leaders are shifting 2026 budgets, where retail media gains are coming disproportionately from brands with challenger-share dynamics, not category leaders.

Building Your 2027 Retail Media Budget

Start the 2027 planning cycle with a channel-by-channel test, not a wholesale shift. The benchmark table above is a starting hypothesis, and the fastest way to turn it into a real budget is a controlled four-week split test across at least three networks before you lock next year's spend.

1. Run a matched-category test on Amazon, Walmart Connect and Instacart simultaneously, holding creative and offer constant, and compare cost per incremental unit sold, not platform-reported ROAS.
2. Request a lift study or clean-room holdout from at least one network before renewal. Kroger, Walmart and Amazon all now offer some version of this; ask specifically for it in writing.
3. Segment CPM tolerance by category economics. A high-margin, low-loyalty SKU can absorb a $20 Kroger CPM more comfortably than a thin-margin staple can.
4. Rebalance quarterly, not annually. Retail media auction dynamics shift with seasonality and competitor entry faster than most annual planning cycles account for.

This is also the moment to fold retail media into your broader martech stack audit for budget season, since most retail networks now require their own measurement tooling, clean room access, or creative specs that your existing DSP or attribution stack may not natively support.

None of this replaces judgment. The networks with the lowest CPMs in our table, Instacart and Walmart, earn that discount partly because they still have less advertiser density than Amazon, and density rises every quarter as more brands read benchmark tables exactly like this one. Lock in the arbitrage now, because it will not last through 2027.