Walmart Retail Media: How Walmart Connect Actually Works

What Walmart retail media actually is

Walmart retail media is the advertising business Walmart runs under the name Walmart Connect. It's not a separate company — it's Walmart selling ad placements against its own first-party shopping data, the same way Amazon sells ads against Amazon's data. If you searched "walmart retail media" or "walmart connect retail media," you're asking about the same thing: a retailer-owned ad network built on real purchase behavior instead of third-party cookies.

Walmart Connect sells five things: sponsored search (ads in search and browse results), onsite display (banner and video placements across the site and app), offsite media (Walmart audience data used to buy ads on other publishers, social and streaming), Brand Shop and Shelf (a free storefront for eligible suppliers and marketplace sellers), and store ads (screens and audio inside physical Walmart locations). Most explainers stop at that list. The harder — and more useful — question is how the auction and the measurement actually work, because that's where advertisers lose money without noticing.

The five ad formats Walmart Connect sells

Here's what each format actually buys you, stripped of marketing language:

How the Sponsored Search auction works: a worked example

Sponsored Search runs as a standard auction: you set a cost-per-click bid, Walmart ranks placements on bid and relevance, and you pay only when someone clicks. Here's the arithmetic, using round numbers to show the mechanics — not a real campaign result:

• Budget: $5,000/month
• Average CPC: $1.20 → roughly 4,166 clicks
• Conversion rate: 8% → roughly 333 orders
• Average order value: $30 → $10,000 in attributed sales
• Attributed ACOS: 50% ($5,000 spend against $10,000 attributed sales)

That 50% ACOS is the number Walmart's dashboard will show. It is not the same as knowing whether those 333 orders would have happened anyway — through organic search, a loyal customer's normal shopping trip, or a store visit with nothing to do with the ad. Walmart's reporting tells you what it can see. It doesn't tell you what would have happened without the ad. That gap is the single most misunderstood part of retail media, on any platform, and it's not unique to Walmart.

What Walmart's measurement can and can't tell you

Walmart Connect's closed-loop measurement ties ad exposure to a purchase using first-party transaction data. That's real, and it's a genuine advantage over media that can't see the register. But closed-loop attribution and incrementality are different claims. Attribution answers "did a purchase happen after exposure." Incrementality answers "did the ad cause the purchase." Walmart's own reporting, like most retail media reporting, answers the first question and gets read as if it answered the second.

We run into the identical problem on Amazon DSP, which is the platform we manage. Across 30 advertisers in July 2026, our book returned 6.04x return on ad spend, 78.4 million impressions at a $4.00 CPM, and a blended $1.42 cost-per-click, with a blended cost-per-acquisition of $5.49 across 57,137 attributed purchases — 20.1% of them net-new to the brand. Every one of those numbers is an attributed number. We only call them incrementality once we've checked them against a holdout: split an audience, one group sees the ad, one doesn't, and the gap between the two groups is what the ad actually added. Walmart Connect doesn't hand most advertisers an equivalent holdout tool today. If you want to know whether your Walmart spend is incremental, you generally have to build the test yourself — a store-cluster split, a DMA split, or a paused-versus-live period — because the platform's dashboard won't do it for you.

The mistakes that cost the most money

The mistakes we see most often, including patterns we've had to correct in our own campaigns on other retail platforms:

• Treating attributed ROAS as proof. A campaign can show a strong attributed return and still have added close to nothing, because the shopper was buying anyway.
• Ignoring Brand Shop and Shelf because it's free. Free doesn't mean low-value — it's often the highest-intent real estate a brand gets, and it sits unbuilt because no one owns it.
• Running Sponsored Search and skipping store ads and offsite entirely. Walmart's real advantage is the closed loop between digital exposure and in-store purchase. A search-only strategy throws that advantage away.
• Scaling spend without checking stock status. An ad driving clicks to a low-stock or out-of-stock item is the fastest way to burn budget with nothing to show for it.
• Reading offsite results the same way as onsite. Offsite campaigns run on someone else's inventory using Walmart's audience data — the attribution logic and reporting lag aren't identical to onsite placements, and blending them hides which one actually worked.

If your numbers look wrong: what to check, and where this connects to Amazon

Work through this order before touching the budget:

• Check the attribution window first. A sudden jump or drop in attributed sales is often a reporting lag, not a real change in performance.
• Check stock status on the advertised SKUs. Clicks landing on an out-of-stock item show up as wasted spend with no obvious cause.
• Separate onsite from offsite performance before drawing conclusions — a blended ROAS hides which format actually moved.
• If a fix "didn't work," check whether it actually shipped. Bid changes and budget shifts inside Ad Center can take a reporting cycle to show up; a flat line the day after a change usually means it hasn't taken effect yet, not that it failed.
• If you suspect the attributed number is inflated, build a holdout before cutting spend. Cutting a campaign on an unproven incrementality assumption throws away the campaigns that were actually working alongside the ones that weren't.

Dr. DSP doesn't run Walmart Connect — it's Amazon DSP, Amazon's programmatic display, video and audio platform, run as a managed product by Fable 5 out of Full Circle, which has managed more than $500M in revenue across 100+ brands. If your brand advertises on both Amazon and Walmart, the incrementality question in this piece applies to your Amazon budget the same way it applies to Walmart's. That's the specific problem Dr. DSP is built to answer, using holdouts and matched controls reconciled in Amazon Marketing Cloud — not a Walmart-specific fix, but the same discipline applied wherever the attributed number and the true lift number might disagree.