Beyond Amazon: How Retail Media Is Reshaping CPG Budget Allocation

Retail media stopped being optional in 2024. For CPG and consumer brands selling through retailers, the question shifted from "should we run retail media" to "how much, where, and at the expense of what other line item."

The honest answer is harder than the marketing decks make it sound. Retail media is genuinely reshaping CPG budget allocation, but the measurement layer is incomplete, the trade-off with traditional trade spend is rarely modeled cleanly, and the retailer networks are not interchangeable. Treating it as a unified channel is the first mistake.

Retail media stopped being optional in 2024

The growth numbers are not subtle. Retail media ad spend in the U.S. crossed $50 billion in 2024 and continues to grow at 15 to 20% annually, faster than every other digital category. Amazon Ads alone is now the third-largest advertising business in the world.

For CPG brands, that growth has come with structural change. Sponsored Products on Amazon used to be a tactical lever; it is now the table-stakes investment for any CPG with meaningful Amazon revenue. Walmart Connect has grown to a multi-billion-dollar business in five years. Target Roundel, Kroger Precision, Instacart, Home Depot, Lowe's, and a dozen others have built networks that are no longer experimental.

The shift is not just budget growth. It is budget moving from places it used to live — traditional digital, trade spend, sometimes brand — into the retail networks. The brands that have not modeled that reallocation are the ones whose total go-to-market spend keeps growing while DTC and brand-side performance flattens.

The category map

The retail media category is not a single channel. It is a fragmented set of networks with different commercial models, measurement standards, and audience compositions.

Amazon Ads is the largest and most mature. Sponsored Products, Sponsored Brands, Sponsored Display, and Amazon DSP each function as a distinct sub-channel. Amazon's measurement is the most developed of any retail network, including post-purchase attribution that no other retailer can match.

Walmart Connect is the closest competitor in scale. Strong in-store integration, growing programmatic capability, and a broader audience profile than Amazon. For CPG brands with meaningful Walmart shelf presence, it is increasingly non-optional.

Target Roundel over-indexes on higher-income shoppers, beauty, home, and apparel. Its first-party data is among the most useful in retail media for audience targeting beyond the immediate purchase context.

Kroger Precision Marketing is the deepest data offering in grocery, with Kroger's loyalty card data anchoring an audience layer that other grocery networks can't replicate.

Instacart Ads is a closed-loop performance channel for grocery and CPG, with strong attribution within the platform but limited off-network reach.

The category networks (Home Depot, Lowe's, Best Buy, Ulta, Sephora, CVS, Walgreens, and others) each serve a specific buyer profile. For brands with meaningful presence in those categories, they are increasingly defensible spend.

These networks are not interchangeable. The same dollar produces different outcomes on each, depending on category fit, shopper data, and the brand's existing retailer relationship.

Why retail media is reshaping CPG specifically

Three structural advantages explain why retail media has captured CPG budget faster than any previous channel.

Closed-loop measurement. Retail networks can attribute ad exposures to actual purchases at the retailer, with first-party data the brand doesn't have access to anywhere else. For CPG, where DTC visibility is limited and most sales happen in stores or through retailers, this is the only category where sales attribution is structurally clean.

Shopper data. Retailer first-party data is more valuable for CPG than for any other category, because the shopper data describes the actual category buyer. Targeting "people who bought a competing product in the last 90 days" is a level of precision DTC channels cannot match.

On-site placement at the point of decision. Retail media reaches shoppers actively comparing products in the category, often with the product page already open. The intent profile is structurally different from any awareness channel.

Combined, these advantages have produced the fastest budget reallocation in modern advertising history. CPG brands that resisted have watched competitors capture share at retail while their own digital programs underperformed.

The trade-off most brands miss: retail media vs. trade spend

The honest conversation about retail media is the trade spend conversation. Retail media is funded — partially or entirely, depending on the brand — from budget that used to go to trade promotion. The reallocation is rarely modeled cleanly.

Trade spend (price discounts, slotting fees, in-store promotions) and retail media are not perfectly substitutable, but they overlap. A dollar spent on Sponsored Products may produce the same incremental lift as a dollar spent on a temporary price reduction — or significantly more, or significantly less, depending on category, retailer, and shopper behavior.

The brands that handle this well treat retail media and trade spend as one combined budget, modeled together, with reallocation governed by measured outcomes. The brands that handle it poorly run them as separate line items with separate teams and separate KPIs, and end up over-funding both.

Sponsored Products vs. Display vs. Off-site media

Within a single retailer network, the sub-channels behave differently and require different evaluation.

Sponsored Products is the high-intent capture layer. Shoppers searching the category, seeing a sponsored result for a relevant SKU. Measurement is direct, attribution is clean, and the math is closer to performance marketing than brand investment. This is where retail media spend tends to anchor.

Display and Sponsored Brands sit higher in the funnel. They build awareness within the retailer's environment, often without driving immediate purchase. They work, but they cannot be evaluated on the same ROAS framework as Sponsored Products. They are awareness investment with a retailer attribution layer attached.

Off-site media (programmatic display, CTV, audio bought through the retailer's media network using their audience data) is the newest and least mature. The promise is reaching the retailer's shopper outside the retailer's own properties, with the retailer's data informing targeting. The measurement is still developing. For brands with the scale to test, the data layer is genuinely differentiated. For brands without, it is often premature.

Budget allocation framework

The right allocation across retail media depends on four variables: the brand's category mix, retailer concentration, DTC strength, and competitive pressure within each network.

For a CPG brand with 30 to 50% of revenue through Amazon, Amazon Ads alone often justifies 8 to 15% of total marketing spend. For a brand with significant Walmart presence, Walmart Connect should grow to a similar magnitude within 12 to 18 months of meaningful investment.

The portfolio question is harder. Most CPG brands need a balanced retail media investment across two to four networks, with weighting based on revenue contribution. Concentrating all retail media on one retailer (usually Amazon) is the most common allocation pattern and usually wrong — it under-invests in retailers where the brand could be growing share.

The brands that do this well treat retail media allocation as an analytical exercise, not a relationship one. The retailer with the loudest sales team is rarely the retailer with the highest incremental ROAS.

The measurement problem

Retail media's measurement story is better than most channels and still incomplete in specific ways.

The strong part: retailer-attributed sales are the closest thing to ground truth in digital advertising. The ad exposure and the purchase happen in the same system, often within hours, often with cart and SKU-level data.

The weak part: retailer-attributed sales over-credit the channel. A shopper who saw a Sponsored Product ad on Amazon may have purchased anyway — driven by brand demand created elsewhere, by category awareness, by the retailer's own promotion. The platform-attributed ROAS is the upper bound, not the truth.

The triangulation that produces an honest read involves incrementality testing within the retail network (most major networks now offer this), modeling retail media within an MMM, and reconciling retailer-attributed numbers against broader brand growth signals.

Brands that take retailer attribution at face value tend to over-invest in retail media. Brands that ignore retailer attribution entirely under-invest. The honest read is the triangulated one.

When retail media is incremental vs. when it isn't

The single most important question in retail media is incrementality: how much of the retailer-attributed revenue would have happened anyway.

Three patterns indicate retail media is doing incremental work: branded search volume rising in the retailer's own environment (Amazon search for brand name growing), new buyer rates climbing within the retailer's loyalty data, and category share growing within the retailer measured independently of the ad spend.

Three patterns indicate retail media is mostly intercepting demand: branded search flat while retail media spend grows, repeat-buyer concentration in the retailer-attributed sales, and parallel performance with DTC where retail media is supposedly outperforming.

The brands that scale retail media intelligently are the ones that test incrementality formally, at least once a year per major network. The ones that don't end up funding what they would have sold anyway.

The Takeaway

Retail media is no longer optional for CPG and most consumer brands selling through retailers. The question is not whether to invest. It is how to allocate across networks, how to handle the trade-off with trade spend, and how to measure incrementally instead of by retailer attribution alone.

The brands handling this well treat retail media as a coordinated portfolio across networks, not as a single channel and not as parallel silos. They reallocate trade dollars deliberately, model incrementality, and read retailer-attributed numbers as the upper bound — not the answer.

If you're working through the retail media portfolio question for your brand, our team can help structure the allocation framework and measurement honestly. Learn more about running retail media as a coordinated channel.