A Practical Guide to Retail Media Networks

Amazon proved a retailer could build a genuine advertising business on top of its own shopper data, and nearly every retailer with meaningful scale has now followed. That’s created a landscape of dozens of separate retail media networks, each with its own login, its own ad formats and its own reporting, and the practical challenge for most brands right now isn’t whether to use retail media. It’s figuring out which of these networks deserve budget.

The networks worth knowing right now

Amazon Advertising remains the largest and most mature network by a wide margin, with the deepest self-serve tooling and the most third-party case studies to learn from. Walmart Connect has grown into one of the fastest-growing parts of Walmart’s business, and the company’s pending acquisition of Vizio, expected to close by the end of this year, signals a clear ambition to add owned connected-TV inventory to its ad offering rather than staying purely on-site. Instacart Ads has spent this year building out self-serve tools and expanding its reach beyond its own app into offsite placements, a meaningful shift for a network that started narrowly focused on its own search results. Target’s Roundel, Kroger Precision Marketing and a growing list of networks from Costco, Chewy and other specialty retailers round out a landscape that keeps adding new entrants nearly every quarter.

Why the fragmentation keeps getting worse

Every retailer with a meaningful base of logged-in shoppers has the same incentive to build a network: advertising is close to pure margin against retail’s typically thin margins, and first-party purchase data has only gotten more valuable as broader browser tracking keeps eroding. That math doesn’t stop being attractive once a handful of retailers have already launched networks, so the number of platforms competing for ad budget keeps growing rather than consolidating. For a brand, that means more separate logins, more separate reporting dashboards and less ability to compare performance across networks on a common basis.

How to choose where to spend

• Start with where your products sell. A network where you have thin distribution or low sales rank will rarely produce good advertising results, no matter how well the platform itself is built.
• Weigh maturity against differentiation. Amazon offers the deepest tools and the most proven playbooks; smaller or newer networks may offer less competition for the same shopper attention and a more specific audience.
• Look at whether the network extends off-site. Walmart’s Vizio deal and Instacart’s offsite expansion both point toward networks using shopper data to reach customers beyond their own store, which changes how you should think about a network’s total reach.
• Resist the urge to be everywhere at once. Two networks run with real product-listing optimization and careful measurement will outperform six run thinly, and most brands are better served proving the model on a couple of platforms before expanding further.

Running any of these networks well

The specific interface differs from network to network, but a few disciplines carry over regardless of which platform you’re in. Your product listing is effectively your landing page on every one of these networks, so titles, images, reviews and in-stock status matter more to conversion than bid adjustments do. Defend your own branded search terms first, since they’re typically cheap and protect existing demand from competitors bidding on your name, then expand into category terms deliberately rather than all at once. And be skeptical of a flattering return on ad spend from any single network in isolation, since it often reflects demand that would have converted anyway rather than growth the ad created.

The year ahead

Expect the number of retail media networks to keep growing before it consolidates, as more retailers with a real base of first-party shopper data recognize the same opportunity Amazon and Walmart already capitalized on. Brands that build a repeatable process for evaluating and onboarding a new network, rather than treating each one as a bespoke project, will handle that continued fragmentation far better than those still figuring it out network by network.