Retail Media Networks: The Ad Channel Quietly Outgrowing Google and Meta
For years, digital advertising basically meant two names: Google and Meta. That's changed faster than most small store owners have noticed. Retail media — ads bought directly through Amazon, Walmart, Instacart, and similar platforms — pulled in roughly $128 billion in advertiser spend in 2026, and it's now the fastest-growing major ad category in US advertising history.
[Umair: Worth watching closely if you're already selling through Amazon or Walmart even a small, well-tracked test budget here can reveal a lot about where your category's ad costs are actually heading.]
If you sell through Amazon, Walmart, or a similar marketplace, this isn't a trend to casually watch from the sidelines — it's increasingly where the advertising budget in your category is actually going, and understanding it is becoming close to a baseline requirement for competing effectively.
What a Retail Media Network Actually Is
A retail media network (RMN) is an advertising platform owned by the retailer itself — Amazon Ads, Walmart Connect, Target's Roundel, Instacart Ads, Kroger Precision Marketing, and a growing list of others. Instead of buying an ad on Google or Instagram and hoping it eventually leads to a sale on a completely separate platform, a retail media ad runs directly inside the store where the purchase actually happens — a sponsored product in search results, a banner on a category page, a screen at checkout.
The reason this has grown so fast comes down to three structural advantages no open-web ad platform can fully replicate:
Real purchase data, not just browsing behavior. A retailer knows exactly what a shopper has actually bought before, not just what they clicked on or lingered over.
Closed-loop measurement. Because the ad and the purchase both happen inside the same platform, the retailer can report with real confidence whether a specific ad actually drove a specific sale — something a lot of open-web advertising has always struggled to prove cleanly.
Reduced dependence on third-party cookies. As browser-level tracking keeps getting restricted, first-party retailer data (built from actual account purchase history) has become a genuinely more reliable targeting foundation than the cookie-based tracking that used to dominate.
Who's Actually Winning This Space
The market is heavily concentrated. Amazon Ads alone is estimated to generate somewhere north of $50 billion in annual ad revenue, commanding roughly 70-80% market share depending on the estimate. Walmart Connect sits a distant second — genuinely smaller in raw dollar terms, but growing quickly and with a real structural advantage Amazon doesn't have: the ability to tie an online ad directly to an in-store purchase through Walmart's loyalty and payment data.
Below those two, a real "second tier" has formed — Instacart Ads, Kroger Precision Marketing, Target's Roundel, DoorDash Ads, Costco Media, and a growing list of category-specific networks (Best Buy for electronics, Ulta and CVS for beauty and health). Analysts expect Amazon and Walmart together to absorb the overwhelming majority of new retail media dollars in 2026, but the smaller, category-specific networks are carving out real value for brands that sell heavily within their specific niche.
Why This Matters Even If You're Not a Huge Brand
It's easy to assume this space is only relevant to large CPG companies with dedicated ad teams. In practice, a lot of what makes retail media effective is genuinely accessible to a smaller seller, particularly on Amazon and increasingly on Walmart:
Sponsored Products ads are the entry point for most sellers — a cost-per-click model that boosts a specific listing's visibility in search results, functionally similar in setup complexity to a basic Google Ads campaign, just running inside the marketplace instead.
The data advantage cuts both ways. A smaller seller doesn't have the massive ad budget of a national brand, but they also benefit from the same closed-loop measurement — you can actually see, with real confidence, whether a specific ad spend led to a specific sale on your own listings, which is more concrete feedback than a lot of open-web advertising ever gives a small seller.
Category-specific networks can offer better value than the giants. If your product fits neatly into a specific vertical — home improvement, beauty, pet supplies — a smaller, more targeted retail media network in that category can sometimes deliver a better cost-per-result than fighting for visibility in Amazon's much more crowded, competitive auction.
The Real Tradeoffs Worth Understanding
Reporting isn't standardized across networks. Amazon might report total sales attributed to a campaign, Walmart Connect reports gross merchandise value, and a grocery-focused network might separate in-store versus online sales entirely. Comparing performance across more than one network isn't a simple like-for-like exercise — it takes real attention to understand what each platform is actually measuring.
High-competition categories can get genuinely expensive. In categories with a lot of established sellers already bidding, cost-per-click on Amazon Sponsored Products can climb well past what a similar click would cost on the open web — it's worth watching your actual return, not just assuming retail media is automatically cheaper because it's "closer to the sale."
You generally need to already be selling on the platform. Most of these networks are built around advertising products that are already listed and fulfillable on that specific marketplace — this isn't a general brand-awareness channel in the way a social ad might be; it's much more tightly tied to an existing catalog and storefront on that retailer.
A Reasonable Way to Start
For a smaller store already selling on Amazon or Walmart, a sensible entry point looks something like this:
Start with Sponsored Products on your best-reviewed, highest-margin listings. These convert more reliably once someone actually sees them, which makes the early testing budget go further while you're still learning how the auction behaves for your specific category.
Set a modest test budget and actually track it against your own numbers, not just the network's own reported metrics — comparing cost-per-click and resulting margin against your actual product economics matters more than any headline benchmark from an industry report.
Expand only after you've seen it convert reliably. Agencies working in this space commonly recommend allocating something like 10-15% of a broader ad budget to testing a newer or smaller retail media network, only shifting more spend over once real performance data justifies it — the same logic applies to a small seller testing Walmart Connect or a category-specific network for the first time.
Don't abandon your other channels while testing this. Retail media is a genuine complement to search and social ads, not a wholesale replacement — the fundamentals of running paid campaigns well largely carry over, even though the specific auction mechanics differ from platform to platform.
Where This Is Headed
The market looks likely to keep consolidating around a small number of large, sophisticated networks (Amazon, Walmart) alongside a growing "long tail" of smaller, more specialized retail media platforms built on shared underlying ad infrastructure rather than each retailer building everything from scratch. For a seller, the practical implication is that this channel isn't a passing trend to revisit later — it's increasingly a standard part of the marketing mix for anyone selling through a major marketplace, alongside search and social.
What to Actually Measure Before Scaling Up
Before committing more budget to any retail media network, a few numbers are worth tracking honestly rather than relying purely on the platform's own dashboard summary:
Advertising Cost of Sales (ACoS) — your ad spend as a percentage of the resulting sales. A campaign with a high ACoS can still be worth running if it's driving new customer acquisition or clearing older inventory, but it's worth knowing that number rather than assuming visibility alone justifies the spend.
Organic rank movement alongside paid spend. On Amazon specifically, a well-performing sponsored campaign can sometimes lift a listing's organic search ranking over time as sales velocity increases — worth checking whether that's actually happening for your listings, since it changes the real return of a campaign beyond the immediate paid clicks.
Repeat purchase rate from ad-driven customers. A customer acquired through a retail media ad who comes back and buys again organically later represents a very different return than one who never purchases again — this is harder to track precisely, but worth estimating where the platform's own reporting allows it.
Common Questions
Do I need a big budget to start with retail media ads? No — Sponsored Products campaigns on Amazon, for instance, can start with a genuinely modest daily budget, similar to a basic Google Ads test. The bigger investment tends to be the time spent learning how bidding and placement actually behave for your specific category.
Is Amazon Ads worth it if I also sell on my own Shopify store? Generally yes, if a meaningful share of your sales already come through Amazon — the ad spend there specifically boosts visibility within that marketplace and doesn't compete with or replace your Shopify-focused marketing efforts.
How is this different from a regular Google or Meta ad? The core difference is where the purchase happens. A Google or Meta ad sends someone to your website to complete a purchase; a retail media ad keeps the entire journey — click and purchase both — inside the same retailer's ecosystem, which is what enables the more precise, closed-loop measurement these networks are known for.
Which retail media network should a small business start with? Generally whichever marketplace you already sell on most actively — starting with Amazon Ads if Amazon is your primary sales channel, or Walmart Connect if that's where more of your existing volume comes from, rather than trying to test several networks simultaneously from a standing start.