Retail Media Networks: When Marketplace Ads Become Rent
By Diosh — Founder, AHAeCommerce | eCommerce decision intelligence for $50K–$5M GMV operators
A brand ranking organically in the top three positions for its primary Amazon category keyword has already paid for that shelf position — through product reviews, listing optimization, sales velocity, and margin sacrificed to build the ranking over 18–24 months. When the same brand then buys Sponsored Product ads to defend that ranking against competitors bidding on the same keyword, it pays twice for the same demand. The first payment built the shelf; the second payment rents it.
U.S. retail media spend reached $60B in 2025 and is forecast to hit $100B by 2028 (Nielsen/eMarketer). That growth is not driven by brands discovering new demand — it is driven largely by existing marketplace sellers bidding for consideration share they already earned organically. The operational question that most retail media management frameworks don't ask is whether the spend is producing incremental revenue or simply buying back exposure the algorithm was previously awarding for free.
The Default Assumption (and Why It Fails)
Retail media is consistently positioned as performance marketing "closer to purchase." The logic: ads served on Amazon, Walmart, Target, or Instacart reach buyers who are already in a shopping context with credit card in hand. Conversion rates are high, attribution is clean, and the return on ad spend metrics are better than upper-funnel channels. This framing is correct as far as it goes — retail media does convert well.
Where it fails is in distinguishing what kind of revenue it's converting. High-converting retail media can be capturing demand you would have captured anyway without the ad spend, from customers who were already going to buy from you. The metric that most retail media reporting shows is ROAS — return on ad spend. The metric it consistently omits is incrementality: what portion of the conversions happened because of the ad, vs. what portion would have occurred through the organic listing.
The distinction is not academic. If 60% of your Sponsored Product conversions would have converted organically, your effective ROAS is not 4x — it's less than 2x after accounting for the displaced organic revenue. Your real media efficiency is worse than it appears, and the spend that looks like acquisition is partially self-cannibalization.
What the Decision Actually Hinges On
Incremental Demand vs. Defensive Spend
Not all retail media serves the same purpose, and conflating purpose with performance metric produces incorrect investment decisions.
Acquisition spend targets keywords, competitor ASINs, or categories where you don't rank organically and where you're genuinely reaching buyers who wouldn't otherwise encounter your product. ROAS for this spend should be evaluated against customer acquisition cost benchmarks, not against blended retail media ROAS. Conversion rates will typically be lower; the relevant measure is whether the acquired customer returns and what the LTV-to-CAC ratio looks like.
Defensive spend protects your branded keywords, your own ASIN listing, and your top organic category positions from competitor conquest. The ROAS metrics on defensive spend look excellent — you're capturing customers who were already looking for you — but the incrementality is low. Without defensive spend on your own branded terms, a competitor's ad appears in the space your organic listing was occupying. Defensive spend is not growth spend; it is the cost of retaining demand you already earned.
Rank maintenance spend occupies a gray zone: ads that maintain sales velocity sufficient to preserve organic ranking position. On Amazon, sales velocity is a ranking input. Pausing ads on a well-ranked listing can cause organic rank decay, which then requires more aggressive ad spend to recover. This creates a cycle where the ad spend is structurally necessary to maintain the organic position that was built on the assumption of continued ad support — the original sin of marketplace dependency.
Understanding which of these three categories your retail media budget falls into is the prerequisite for any meaningful spend optimization.
The Margin Stack at the Product Level
Retail media spend on marketplace channels sits on top of an already-compressed margin stack. For a typical Amazon FBA seller:
The gross margin per unit before any advertising is approximately: Selling price minus product cost minus FBA fulfillment fee ($3.45–$5.55 for standard-size items) minus referral fee (typically 15%) minus inbound shipping minus storage. For a $50 product with a $15 landed cost, a 15% referral fee, $4.50 FBA fee, and $0.50 in storage, the pre-ad gross is approximately $13.50, or 27% gross margin.
Adding Sponsored Product spend at a 10% advertising cost of sale (ACoS) — which is a common target — reduces the per-unit contribution to $8.50, or 17%. Adding the defensive branded keyword spend, which often runs another 3–5% ACoS on branded terms, reduces it further.
At 15% combined ACoS on a 27% pre-ad margin, the product is operating at 12% net contribution before seller central fees, returns, and any off-platform marketing cost. The product that appeared profitable at the gross margin line is a marginal contributor after retail media is layered in. The ROAS at the ad level was 4x. The margin story at the product level looks very different.
When the Platform Controls the Bid Floor
As retail media auctions mature, minimum viable bid floors rise. A category where the top-of-search sponsored placement cleared at $0.45 cost-per-click in 2022 may clear at $1.20 or more in 2025 as more brands enter the auction. For brands with products where the pre-ad margin is already thin, bid floor inflation compresses contribution margin proportionally with each auction cycle.
This is the mechanism by which marketplace advertising becomes rent: the floor rises because more bidders compete for the same shelf, not because the underlying demand has grown. An operator whose retail media costs increased from 8% to 14% ACoS over two years, while organic rank held steady and conversion rate was stable, is experiencing bid inflation, not increased media value. The additional 6% is being extracted by the platform auction.
The Cost Reality
The clearest way to see retail media rent extraction is to build the full contribution margin model with and without retail media:
| Without retail media | With retail media at 12% ACoS | |---|---| | Gross margin: 27% | Gross margin: 27% | | Retail media cost: 0% | Retail media cost: 12% | | Contribution margin: 27% | Contribution margin: 15% | | Organic sales volume: 100 units (example) | Ad-assisted sales volume: 140 units | | Total gross profit: 27 units × margin | 140 units × 15% margin |
The math shows whether incremental volume at 15% margin is worth more than the foregone volume at 27% margin. For products where the 40% volume increase at the lower margin generates more total gross profit, retail media is value-creating. For products where the incremental volume lift is modest relative to the margin compression, retail media is rent.
Retail media platforms do not show you this model. They show ROAS, ACoS, CTR, and conversion rate. None of these metrics answer the question: "Am I making more total gross profit with this spend than without it?"
The Trade-Off Map
Spend and Win Volume at Lower Margin
The upside: Retail media can buy market share faster than organic rank builds it. In competitive categories, paid consideration can hold or extend category position during periods when organic rank would otherwise erode. For products with strong unit economics that hold contribution profitability even after 12–15% ACoS, the volume gain compounds in review accumulation and sales velocity — both of which build long-term organic position.
The downside: In categories where organic rank is determined primarily by conversion rate and review quality — not by sales velocity alone — retail media spend produces category share at the cost of contribution margin without meaningfully strengthening the organic position. The volume buys awareness; it doesn't build a compounding moat.
Pull Back and Protect Margin
The upside: Eliminating low-incrementality retail media spend recovers the margin compression immediately. For products where organic rank is stable, the recovery can be substantial — 8–12 percentage points of contribution margin restored, with modest volume impact if incrementality is truly low.
The downside: Pulling back spend completely on competitive category keywords, even when incrementality is questioned, creates an exposure window. Competitors fill the sponsored placement vacated. Organic ranking doesn't move in the first two weeks; in some categories it erodes over 30–60 days as click-through traffic decreases. The transition from paid to organic-only is not seamless.
The Incrementality Measurement Approach
The most useful intermediate path: run incrementality measurement before making hold/cut/increase decisions. Amazon offers holdout testing at the campaign level; Walmart Connect has similar functionality. A 2–4 week holdout test on non-branded category spend, during a period without major promotions, reveals the true incrementality rate and allows accurate ROAS calculation.
If incrementality testing shows 30% of conversions are incremental (70% would have occurred organically), the effective ROAS on incremental revenue is one-third of reported ROAS. That recalculation often produces the decision immediately.
When to Expand vs. When to Hold
Expand retail media spend when: you are entering a category keyword where you have no organic visibility, where the target customer is demonstrably different from your current buyers, and where you have modeled contribution margin at the expected ACoS floor. These conditions describe genuine acquisition.
Defend at current levels when: branded keyword spend is holding stable ROAS, organic rank is at target, and reducing spend triggers rank decay within 30 days. This is the documented rent cycle — you've established that the organic position requires ad support to sustain.
Reduce or eliminate when: holdout testing shows incrementality below 25%, organic rank holds within 5% of target during test period, and the spend is concentrated on top-of-funnel category keywords where your organic listing is already in positions 1–5. These are the spend categories most likely to be capturing existing demand at a margin cost.
What Operators Get Wrong Most Often
The most common mistake is optimizing retail media for ROAS without measuring incrementality. ROAS is a ratio of revenue to spend. It doesn't tell you whether that revenue would have occurred without the spend. A 5x ROAS that is 80% non-incremental is less efficient than a 2x ROAS that is 90% incremental, because the first case is spending $1 to capture $4 of demand you already owned.
The second mistake is treating retail media as a budget line separate from the marketplace margin model. Operators who evaluate retail media performance in the marketing dashboard and marketplace contribution margin in the P&L rarely see the full picture simultaneously. Retail media spend reduces the effective contribution margin from the marketplace channel. The evaluation belongs in the same model.
The third mistake is allowing bid floors to auto-escalate without an ACoS ceiling. Bid management tools that optimize for conversion rate without an ACoS constraint will bid up to clear the auction, regardless of whether the resulting margin makes economic sense. Explicitly setting contribution-margin-based ACoS ceilings — not ROAS targets — ensures that ad optimization isn't trading margin for metric performance.
Retail media is not inherently rent, and it is not inherently growth. It is both at different times in the same account, sometimes in the same campaign. The operators who know which category each spend dollar falls into are the ones who can grow profitably through the channel; the ones who don't tend to discover the problem at the annual P&L review rather than the monthly campaign report.
Build a contribution margin model with retail media included. Run incrementality measurement on your highest-spend campaigns. Know the number. Then decide.
AHAeCommerce is an independent eCommerce decision intelligence platform. No affiliate relationship influences this analysis. Drafted with AI assistance. Edited and claim-tested by Diosh.
Sources: Nielsen/eMarketer, "Future of Retail Media" — nielsen.com/insights/2025/future-retail-media/; Amazon FBA fee schedule, 2024 — see also: Marketplace Fees: What You're Actually Paying, Amazon vs Your Own Store
