多数零售媒体网络本不该存在:它们应如何自证价值
Most Retail Media Networks Shouldn’t Exist. Here’s How They Can Prove Their Worth
零售媒体网络面临严重过饱和与整合浪潮,腰尾部网络必须依靠差异化度量和生态互操作性来证明自身价值。数据显示 Amazon 和 Walmart 在 2025 年拿下了美国 85.5% 的零售媒体支出,其余中小型网络因受众重叠且缺乏跨平台归因透明度,难以分得实质预算。文章指出零售商应打破封闭花园,通过接入 API、统一归因标准并整合 The Trade Desk 等主流采买工具来降低投放门槛。
原文梳理了零售媒体网络的饱和困境与整合趋势,并为腰尾部平台如何通过度量差异化与工具互通争取预算提供了实操思路。
Monday, November 16th, 2026 – 12:35 am
SHARE:

Paul Evans
Founder & Chief Positioning Engineer
We’re well past the point of oversaturation in ad tech.
There are more than 175,000 software vendors listed on G2, over 15,000 marketing technology solutions and reportedly over 12,000 ad tech vendors in operation. Competitive intensity is real.
But no tech market is more ripe for consolidation than retail media. Look no further than DoorDash’s recent $17 billion acquisition of Instacart.
And then there’s Kevel, which supplies ad technology infrastructure to retail media networks and recently raised $52 million. Its founder and CEO, James Avery, told AdExchanger that a lot of retail media companies are selling to the same customers and that “consolidation is going to happen.”
He also said Kevel likes to joke that “we’re going to help launch 1,000 walled gardens, but then the brands are, like, ‘No, please don’t!’”
The brands are right. For many, the media network model is the wrong bet to make.
Differentiation and measurement are real advantages
Amazon and Walmart took 85.5% of US retail media spend in 2025, leaving only 14.5% for every other network, combined. EMarketer expects this duopoly to take 89% of the growth in 2026 and the share held by every other network it tracks to stay flat or fall through 2027.
Brands haven’t stopped adding networks to their plans, though. McKinsey found that a third of US advertisers used nine or more commerce media networks in 2025. But the reality is that a network can appear on a media plan and still receive very little of the budget.
The same McKinsey survey found that as advertisers add networks, their audiences overlap and the networks become harder to pull apart. Ad buyers told eMarketer much the same thing: They struggle to distinguish one platform from another. The retailer’s brand just isn’t enough of a differentiator.
Too many networks exist without addressing the biggest question: “Why should I be chosen?” Most networks are selling to the same brands and addressing the same audiences, while describing themselves with the same claims about first-party purchasing data and closed-loop measurement. As such, buyers have nothing to base their choice on except for scale and price. And no one beats Amazon there.
The untapped opportunity is for retailers to differentiate their measurement stacks.
In NIQ’s CMO Outlook survey, 67% of CMOs said they planned to increase spending on retail media networks in 2026. Yet only 53% believe today’s networks provide adequate measurement and attribution.
Retailers want to see why each decision in an auction was made: “It’s not enough to say, ‘Trust us, this is doing the best it can for you,’” Kevel’s Avery told AdExchanger.
Advertisers are asking for the same evidence. A brand that can’t compare results across networks will most likely keep its budget where the results are easiest to read.
Interoperability becomes a superpower
A retailer’s own shopper data is the one asset no other network can copy. It’s proprietary and unique by definition. But siloing this advantage behind a separate walled login is an expensive and high-friction way of realizing value.
Some retailers have started selling shopper data through partnerships and tools that brands can more easily access.
Tesco, B&Q, John Lewis and Waitrose are piloting dunnhumby’s network alliance, which gives brands a single way to run a campaign across several retailers, with shared standards for measurement and reporting.
Kevel has built API access with Pacvue and Skai along with an integration that lets brands buying Dollar General’s off-site ads through The Trade Desk buy its on-site ads the same way.
In both cases, the retailer still runs its own media business, but brands don’t have to learn its system to buy from it.
Planning for 2027
Avery expects consolidation among the infrastructure vendors that sell to and support retailers, and Kevel now has the money to win that fight. But advertisers will inevitably consolidate their media plans as well. WARC forecasts that, outside Amazon, global retail media growth will slow to 9.8% in 2027. That’s the lowest rate since it began tracking the market. When growth slows, brands will remove the networks they can’t tell apart or can’t compare on performance.
In 2027, every retail media network will have to answer three questions from the advertisers and agencies it sells to. Why should we choose you over Amazon? Can you prove what our media investment did and compare that to other networks? Can we use familiar tools to buy on your network?
Those who can answer all three questions stand the best chance of being part of the media plan next year.
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Follow Paul Evans and AdExchanger on LinkedIn.
来源:AdExchanger · adexchanger.com