Why we stopped recommending Walmart Connect for sub-$5M brands

Two years ago we were Walmart-positive across the board, every Amazon brand we onboarded got a Walmart-channel recommendation in the 90-day plan. We’ve revised. For brands under $5M in Amazon GMV, Walmart Connect is no longer a defensible investment in 2026. Above $5M, the math still works. Below it, you’re funding a learning curve that the platform is no longer subsidizing.

This isn’t a Walmart-vs-Amazon take. We still run multi-marketplace programs. The point is sequencing.

What changed

When Walmart Connect launched in 2021, the auction was deliberately undermonetized, Walmart was buying advertiser participation at a discount to get the program seeded. CPC on a competitive Walmart category in 2022 was 30-50% below the equivalent Amazon CPC. You could buy meaningful impression share at CPCs Amazon charged in 2017. That arbitrage was the real argument for being on Walmart early; the channel-diversification benefit was a secondary, more abstract pitch.

That arbitrage is gone. Median CPC across our Walmart-managed accounts is up 41% YoY. Conversion rates are flat. The math that made Walmart attractive for a sub-$5M brand was cheap traffic, and the cheap traffic is gone. Walmart’s ad revenue is now competitive with Amazon’s on a CPC basis, but with smaller volume per query and worse buyer-intent signal at the long-tail (because Walmart’s search behavior skews more toward category browsing than direct-product queries).

Three other things changed simultaneously, and each compounds the first:

• Walmart’s catalog requirements tightened. The image spec is stricter, the attribute completeness is enforced more rigidly, and SKUs that pass Amazon’s standards routinely suppress on Walmart for compliance reasons that take 2-3 weeks to resolve.
• WFS (Walmart Fulfillment Services) economics shifted. The 2025 fee schedule narrowed the gap with FBA on most categories, removing what used to be a structural advantage for Walmart-fulfilled inventory.
• Walmart’s algorithm is starting to show the same query-rewrite dynamics we see on Amazon. That means the “Cosmo-aware catalog” work you’d do for Amazon now has a Walmart equivalent, and it’s a separate body of work.

Why $5M is the threshold

Below $5M of Amazon GMV, brands typically can’t afford the parallel-investment cost. We costed it on 18 brands last year. Total parallel investment to run Walmart credibly:

That’s $42-68K in the first 90 days. For a $3M Amazon brand, that’s 1-2% of revenue going to a channel that returns less than reinvesting the same dollars in your Amazon program. We modeled the counterfactual on every account we audited, a $40K incremental Amazon DSP test outperformed the same $40K spent on a Walmart launch in 16 of 18 cases.

The two cases where Walmart won were both in narrow conditions: one was a regional CPG brand whose buyer overlap with walmart.com was structural, and the other was a kids’ apparel brand with retail relationships at brick-and-mortar Walmart that the marketplace presence reinforced. Neither had the conditions a generic sub-$5M brand has.

Above $5M, when Walmart works

The math inverts past $5M because the parallel-investment cost stays roughly fixed while your absorbing capacity scales. A $20M Amazon brand spending $50K on a Walmart launch is allocating 0.25% of revenue to a channel that historically returns 8-12% of incremental GMV at 18-24 months. That’s defensible.

The conditions where it works best:

• Your category has structural Walmart-side traffic that doesn’t exist on Amazon (grocery-adjacent, kids’, regional CPG, certain home goods)
• You have a 3PL or WFS-ready inventory model, not a single-FBA-pool model that would need to fork inventory
• You can spare 1-2 ops headcount for the first 90 days, either internal or agency-side
• You already won at Amazon, you’re not running to Walmart because Amazon’s broken

The last condition is the one founders most often violate. Walmart isn’t a fallback for an Amazon program that isn’t working. It’s a multiplier for one that is.

The exception

There is one, branded categories with structural traffic on walmart.com that doesn’t exist on Amazon. The most obvious: kid’s apparel, value-tier home goods, certain regional CPG segments, and grocery-adjacent SKUs that benefit from walmart.com’s “complete the basket” merchandising. If you’re in one of those, the sequencing is different. The retail-trip overlap with walmart.com matters more than the auction efficiency, and the parallel-investment cost is partially offset by Walmart-side acquisition that wouldn’t exist on Amazon at any price.

How to know if you’re ready

Three checks, in order:

1. Is your Amazon program clean? If your Amazon TACoS is rising, your A+ hasn’t been touched in 6+ months, or your inventory health is below 80% on the IPI tracker, fix that first. Walmart will not save you. It will dilute what attention you have left. The cleanest version of this test: run our audit on the Amazon side and resolve every Critical-tier finding before you add a marketplace. If the Amazon side has nothing Critical, you’re a candidate for Walmart sequencing.

2. Can you allocate $50-70K to a 90-day test without it hurting the Amazon program? The capital test isn’t just “can you afford it”, it’s “can you afford it without pulling spend from Amazon.” If Walmart launch funds come from cutting Amazon DSP, you’ve made yourself worse. Pass.

3. Do you have ops capacity? Even with an agency running it, the brand-side ops load on a Walmart launch is 8-12 hours/week for the first 60 days. Catalog approval rounds, Category Manager calls, creative reviews. If you don’t have that bandwidth without delaying Amazon work, the timing is wrong.

What we recommend instead

Stay focused. Treat your catalog like a product. Push DSP. Refine the bps argument on your existing program. Cross $5M cleanly, not on promotion-led volume. Then re-evaluate Walmart in the context of the broader retail-media landscape, not as a checkbox.

The reason we stopped recommending it isn’t that the channel is bad. It’s that $40K in your Amazon program returns more than $40K in a Walmart launch, and that gap isn’t closing in 2026.

Get an audit, we’ll model the Walmart counterfactual on your specific account before you commit a dollar.