Retail Media Networks Media Planning Checklist for CPG / FMCG Brands

Your category's shopper base is not evenly distributed across retailers. Use IRI/Circana or Nielsen POS data, combined with each retailer's sales velocity for your category, to build a retailer index that guides budget allocation. A beverage brand may find that 40% of its volume flows through Walmart — that retailer should command 40% of retail media investment as the starting allocation before optimization adjustments.

RMN investment is increasingly funded from two separate budget pools — brand media (from marketing) and trade/shopper marketing (from sales). On-site sponsored search is often classified as trade while off-site DSP/programmatic is classified as brand media. Agree on the funding source and P&L ownership for each RMN product type before campaign activation, or you'll face retroactive budget conflicts with the sales organization.

Amazon absorbs the majority of CPG retail media budgets because it offers the broadest product category reach, closed-loop attribution, and the largest shopper data asset. But Amazon over-indexes on online-only shoppers. Kroger Precision Marketing, Target Roundel, and Albertsons Media Collective reach the physical grocery shopper who represents 80%+ of CPG volume in most food and personal care categories. Diversify beyond Amazon for grocery-dependent brands.

Most RMN algorithms require a minimum volume of clicks and purchases to optimize effectively. Amazon Sponsored Products needs 30+ purchases/month per campaign to exit learning phase. Spreading $200K across 12 different RMN platforms in small increments produces underperforming campaigns on every platform. Concentrate budget on 3–5 priority RMNs where minimum effective investment can be met, and expand as budgets scale.

For category-leading CPG brands, retail media investment is increasingly linked to joint business plan commitments with retailers — media spend that supports a retailer's sales goals can influence shelf space allocation, end-cap access, and promotional feature frequency. Involve the category management and sales teams in RMN budget planning discussions; media and trade are no longer separate conversations.

Each RMN uses different attribution windows (7-day vs. 14-day vs. 28-day), different ROAS definitions (attributed vs. total), and different 'new-to-brand' metrics. Before comparing performance across Amazon, Walmart, and Kroger, build a normalization framework. A ROAS of 4.0 on Kroger's 30-day attribution window is not comparable to a 4.0 ROAS on Amazon's 14-day window without adjusting for the measurement difference.

For beverages, snacks, personal care, and household staples, Instacart represents a high-intent purchase environment where the shopper is in active cart-building mode. Instacart Ads CPCs are typically $0.35–$0.95 — materially lower than Amazon ($1.21 average) for comparable intent levels. Instacart's Carrot Ads also allow off-site programmatic targeting using Instacart's first-party shopper data.

Sponsored Products on branded ASINs and category keywords form the non-negotiable floor of Amazon retail media. Without brand-term Sponsored Products, competitors conquest your brand searches at low cost. Category keywords drive discovery among non-brand shoppers. Run these always-on with Target CPA or Target ROAS bidding, sizing the budget to achieve 80%+ impression share on branded terms before allocating to expansion.

Sponsored Brand campaigns (banner ads at the top of search results featuring your brand logo, headline, and 3 ASINs) direct to an Amazon Brand Store. The Store provides a brand-owned retail environment within Amazon where you control the product range, content, and cross-sell logic. Brands with active Brand Store campaigns typically see 10–15% higher conversion rates from Sponsored Brand traffic vs. ASIN-direct landing.

Amazon reports 'new-to-brand' purchase percentage at the campaign level — the proportion of purchases from customers who haven't bought your brand on Amazon in the past 12 months. ROAS-optimized campaigns tend to over-index on repeat buyers (higher ROAS, zero incrementality). Establish separate campaign objectives: defend ROAS target on existing buyer keywords; accept lower ROAS on category expansion keywords to optimize for new-to-brand acquisition.

Amazon Sponsored Display allows competitors to appear on your product detail pages under the 'Similar items you might like' and 'Sponsored' sections. Protect your PDPs by activating defensive Sponsored Display on your own ASINs and adding competitor brand terms as negative keywords in your Sponsored Products. Review your own ASIN pages monthly for competitor conquest activity.

CPG brands on Amazon commonly have hero SKUs (top-selling sizes and flavors) and tail SKUs (variety flavors, seasonal items, multipack variants). Budget and bid optimization should prioritize hero SKUs — these generate the highest review volume, organic rank signals, and conversion rates. Tail SKUs often require lower CPCs and more patient ROAS targets to drive new-to-brand trial.

Amazon and Kroger both offer holdout test frameworks where a matched audience group is excluded from ad exposure and compared against the exposed group on purchase behavior. Without holdout testing, ROAS figures are systematically inflated by baseline purchasers (shoppers who would have bought without seeing the ad). Most CPG brands discover that their 'true' incremental ROAS is 30–50% lower than the platform-reported ROAS — and then right-size budgets accordingly.

Retail media's closed-loop attribution is structurally incompatible with most legacy media mix models (MMM) built around TV and digital display. Ensure your MMM provider has a specific retail media input variable — not just 'digital display' — with RMN data fed at the DMA or retailer level. RMNs that are absent from the MMM appear to have zero brand contribution, systematically undervaluing their role in the media mix.

Share of sponsored search voice (SOV) — the percentage of total category keyword impressions your brand captures on Amazon, Walmart, and Instacart — is a leading indicator of shelf health in the digital retail environment. A brand with 30% category SOV losing it to a challenger at 45% SOV is facing a structural distribution problem, not just a ROAS problem. Pull SOV reports quarterly from Amazon Brand Analytics.

ROAS spikes during promotional periods (price reductions, coupons, primary sell) are not evidence of media effectiveness — they reflect price elasticity, not advertising impact. Normalize RMN reporting to exclude weeks with active price promotions when assessing media efficiency. This requires coordinating with the sales and trade marketing team to maintain a promotional calendar that flags promotion dates in campaign reports.

Amazon's private label brands (Amazon Basics, Presto, 365 Whole Foods) and other retailer private labels actively bid on national CPG brand terms. Pull Search Term Reports weekly to identify private label brand names appearing in your keyword match data and add them as phrase-match negatives. Monitor competitor ASIN pages of retailer private labels for conquest ad placement against your branded segments.

Categories where private label has captured 25–30%+ share require proactive 'brand vs. private label' value justification in sponsored ad copy. Creative that calls out formulation differentiation, sustainability credentials, or taste test data performs better in these categories than generic brand-name awareness creative. Amazon A+ Content pages are the ideal surface for this comparative brand story.

ROAS measures media efficiency; market share velocity measures competitive position. For CPG brands, weekly unit share within a category at a specific retailer (tracked via Circana or Nielsen POS) tells you whether your retail media investment is growing your share or simply defending it. Brands that improve ROAS by 0.5 points while losing 2 points of category share are optimizing the wrong metric.

Amazon uses 7-day and 14-day attribution windows; Kroger Precision Marketing uses 30-day; Target Roundel uses 14-day. Comparing ROAS across these platforms without normalizing for attribution window differences produces meaningless cross-platform comparisons. Build a normalization factor for each RMN (dividing the reported ROAS by the window length ratio) before presenting cross-platform performance in any stakeholder report.