The Retail Media Business Case: How Enterprise Retailers Should Model Margin, Not Just Revenue

Last updated: October 2026. Reviewed by Kunal Damgude, Growth and Product Marketing Manager.

Enterprise retailers should build the retail media business case on margin per dollar: the share of each dollar of ad revenue left after the program's own costs, set against the gross and operating margin the retailer earns on a dollar of merchandise sales. An ad placement carries no cost of goods, so the deductions from an ad dollar are platform fees, the ad operations and ad sales teams, measurement, advertiser incentives and, on offsite campaigns, the media bought on other platforms. Finance can accept the case once revenue is sized as a share of gross merchandise value (GMV) by category and phased month by month, starting with onsite sponsored listings. The model also has to carry ad ops effort, advertiser churn and the time brands take to adopt self-serve.

Osmos is a white-label retail media operating system for retailers and marketplaces. Its modular apps for onsite, offsite and in-store ads sit over your existing stack, so you can break your monetisation ceiling faster without rebuilding it.

Brands judge the same spend by return on ad spend (ROAS), the sales they attribute to each dollar of advertising, and ROAS varies by platform, ad format and product category. A retailer's business case needs ROAS to forecast which brands will renew, and it needs margin per dollar to show what the program adds to operating profit.

Why topline ad revenue is the wrong number to lead with

Gross ad revenue overstates what the program earns, because part of it goes straight back out: offsite campaigns include the cost of media bought on other platforms, advertisers receive credits and make-goods when campaigns under-deliver, and a platform paid on revenue takes its share.

Aswath Damodaran, a professor of finance at New York University's Stern School of Business, publishes margins by sector for US listed companies. In his January 2026 data, grocery and food retailers earned a 26.31% gross margin and a 2.29% pre-tax operating margin, general retailers 33.18% and 6.80%, and specialty retailers (the dataset's "special lines" group) 35.30% and 7.73% (NYU Stern, January 2026). A merchandise dollar carries the cost of the goods, and the gross margin left over pays for stores, staff, logistics and marketing before any operating profit remains.

Ad inventory comes with no cost of goods, freight or shrink, which is why retail media can be a high-margin revenue stream: whatever the program keeps after its own running costs adds to operating profit. At the grocery sector's 2.29% operating margin, a retailer needs about $44 of merchandise sales to earn one dollar of operating profit, and at the general retail figure of 6.80% it needs about $15. Finance teams can therefore express the program's contribution as margin-equivalent sales: the merchandise revenue the retailer would need to earn the same operating profit. Each dollar the program keeps equals the operating profit on about $44 of grocery sales, which a topline figure does not show. At Target, ad revenue already shows up in the reported margin rate: in its fiscal 2026 first quarter, Target reported an 80-basis-point improvement in gross margin to 29%, citing growth in advertising and other non-merchandise revenues alongside supply-chain productivity and lower markdowns (Zacks Equity Research on Yahoo Finance, July 2026).

Offsite revenue needs its own line in the model. On Osmos's White Label DSP (demand-side platform), offsite campaigns run on the retailer's own channel seats, meaning its accounts on each outside platform. The retailer sets the audience markup and channel-level fees, so its margin on an offsite dollar is what it sets on top of the media cost.

Margin per dollar is gross ad revenue minus the media passed through on offsite campaigns, advertiser credits, make-goods and incentives, the platform fee, and the cost of the people who run ad operations, ad sales and measurement. The pricing model, whether cost per click, cost per thousand impressions or cost per action, changes where some of those costs fall, for example by adding make-goods when the retailer guarantees impressions. Once the program is running, the margin moves with operating choices such as self-serve campaign control and stronger attribution, which cut overhead and keep advertisers investing when budgets tighten.

The cost side finance will ask about

The cost lines grow at different rates. A platform fee charged as a share of ad revenue rises with revenue, as does the media bought for offsite campaigns. Ad operations and account management grow with the number of advertisers and campaigns, and most integration work is done before launch.

Ramp time, the months before contribution covers the launch spend, needs its own view in the model, because platform, integration and team costs start before the first campaign runs and revenue builds only as advertisers come on board. Count every cost from the month it starts: the payback month is the first month cumulative contribution turns positive, and it becomes the date the board measures the program against. The staffing model and the order in which formats launch both move the payback month.

Headcount enters the model as roles, cost per role and the month each role is added, and ad sales roles belong in it where the retailer goes direct to brands.

Capturing maximum advertiser budget: self-serve and formats for awareness, consideration and performance

Brands fund advertising from several budgets. Performance money usually sits with ecommerce and shopper marketing teams, and awareness and consideration money sits with brand and media teams, so the formats a retailer offers decide which of those budgets it can compete for. Sponsored listings in search results and on category and product pages serve performance budgets, and because the product listing itself becomes the ad, the brand supplies no creative. Placements that put brand creative and product links on the pages where shoppers compare options serve consideration. Homepage and browse banners, video and custom high-impact formats reach shoppers before they search and draw awareness budgets, and each of them runs on creative the brand uploads. A custom format that exists only on the retailer's site can carry a premium price. In the model, each budget type becomes its own revenue line, and formats that run on brand creative add creative-check hours to ad ops.

A sales team covers the largest brands and agencies. Torso and tail brands each have a small budget, and there are too many of them to reach one by one. Self-serve lets those brands run their own campaigns, which brings their combined budget within reach. Osmos's Brand Self-Serve puts brand-facing ad buying on top of the retailer's existing ad server, with briefs, budgets, creative, pacing and reporting in one platform.

India's leading online pharmacy, an Osmos customer, opened new homepage inventory with widgets such as "Skincare Picks", "Cold and Cough Products" and "In the Spotlight", and advertiser demand rose. It generated $484K in ad revenue in July 2025 and recorded a 20% uplift in ad revenue from the homepage after the widget launch. Its budget utilization, the share of their set-aside budgets that advertisers actually spent, rose 26% between July and August 2025. Over the same period the retailer raised keyword-level floor CPCs, so the gain reflects both new inventory and stronger competition for premium keywords.

Kroger's CEO, Greg Foran, described the same kind of change when the company reported its second quarter: "Stronger collaboration between our merchandising and media teams expanded advertising inventory, and optimization efforts improved visibility and conversion for our brand partners" (Digiday, September 2026).

Benchmarks for sizing the opportunity as a share of GMV, without overpromising

Monetization as a percentage of GMV, or M%G, is annual retail media revenue divided by the gross merchandise value that passes through the retailer's channels. As a rate, it lets finance size the program from the retailer's own sales base and compare it across categories, channels and years. The average M%G across Osmos customers is 6.75%. For a first-year plan, build the figure from the bottom up, category by category, and use the average as a reference for where the program can head as its advertiser base and format range grow.

How well a category monetizes depends on the share of its sales that comes from branded products (own-brand lines bring no outside advertiser), the number of brands competing for the same shelf and search results, how much shoppers search and browse the category online, and the size of the marketing budgets that brands in the category run. Finance can check the first three from the retailer's own data.

For a grocer, those drivers put categories with many competing national brands, such as snacks, beverages and personal care, above fresh produce, where much of the range is unbranded. Non-endemic advertisers, such as banks, telecom companies and travel brands, pay to reach the retailer's shoppers without having products on its shelves, so their revenue belongs on its own line outside the category model.

M%G can be built from supply (available impressions, fill rate, price) and from demand (active advertisers, spend per advertiser, renewal share); build both and reconcile. Fill rate is the share of available ad impressions that carry a paid ad. The retailer can track all of these inputs from launch. Sizing each category with a low, base and high value for those inputs keeps the headline number tied to assumptions finance can test, and shows the board which assumption moves the result most.

What the first 12 months typically look like

In the first year, revenue arrives format by format, starting with onsite sponsored listings. They ramp fastest because they run on the search, category and product pages the retailer already operates, and brands' ecommerce teams can buy them on cost per click as soon as the catalog feed is live. Osmos's Product Ads, its sponsored listings format, pick products from the catalog automatically so an advertiser can launch a campaign in one click.

Display, video and custom formats come next, because they need brand creative and draw on awareness budgets that brands often commit during annual planning. Offsite and in-store ramp slowest: offsite needs audience segments, seats on outside platforms and closed-loop measurement in place before brands commit media budgets to it, and in-store needs screens, installation and proof that each ad played, rolled out store by store. In the model, sponsored listings carry the early months, other onsite formats add revenue as brands book their budgets, and offsite and in-store enter late in the first year or in the second.

eMarketer data cited in an Advertising Week column has US off-site retail media spending growing at twice the rate of on-site in 2026. That shows where brand budgets are moving, not how fast a new offsite line will ramp. Build the offsite slope from the retailer's own inputs: audience segments live, brands onboarded to the channel and spend per brand.

First-year targets work best on early inputs, such as advertisers activated and the share that run a second campaign, because those move within the first two quarters, well before a full-year revenue figure exists.

Where the business case breaks

Business cases break when the revenue model leaves out ad ops effort, advertiser churn or the pace of self-serve adoption.

Ad ops effort grows with the number of campaigns: every campaign needs setup, creative checks, pacing and reporting, and a torso or tail brand's campaign needs much of the same work as a large brand's for a fraction of the spend. A model that adds long-tail advertisers to grow revenue while holding ops cost flat overstates margin per dollar. Carry ops hours per campaign by advertiser tier in the model, along with the share of campaigns that run without manual work.

Osmos's Sofie AI Agents take on much of that work for torso and tail advertisers. They set up campaigns automatically, resolve issues in one click, and automate bidding and reporting, so a brand with a small budget can run its campaigns without the retailer's ad ops team handling each step. As a result, retailers on Osmos get higher advertiser adoption among torso and tail brands, and the model carries fewer ops hours per campaign for those tiers.

Year-two revenue depends on advertiser renewals, and brands renew with the retailers that can show them results. NielsenIQ's July 2026 analysis put the share of chief marketing officers (CMOs) planning to raise retail media investment in 2026 at 67%, and the share who consider their networks' measurement and attribution adequate at 53% (NielsenIQ, July 2026).

Model retention by cohort and advertiser tier, test whether brands that joined on launch incentives renew at the same rate as the rest, and tie the retention assumption to the proof of performance the retailer can show at renewal.

Self-serve adds advertisers over months, as brands move from registering to a first campaign and then to repeat spend, so a case that counts every registered brand as a spending advertiser from launch overstates first-year revenue. Model self-serve as that funnel and draw the conversion at each step from the retailer's own early data once it exists. Keep the managed accounts the sales team runs on a separate line.

Before the case goes to the board, run the model with slower self-serve conversion, lower renewal and more ops hours per campaign at the same time, and check that cumulative contribution still turns positive within the plan period.

Framing retail media to the board as a margin lever, and the evidence that lands

A retail media profit and loss statement (P&L) with its own revenue, direct costs, contribution margin and margin-equivalent sales presents the program to the board as a source of operating profit, with an owner who answers for its margin. When ad revenue and its costs are reported inside the marketing budget, the board sees platform fees and headcount first and treats the program as marketing spend.

Ryan Mason, president and chief operating officer of the marketing agency Markacy, told Digiday that "marketing is being evaluated as a financial lever, not just a brand function," and that the questions finance chiefs ask have changed with it: "Conversations that once centered on reach and engagement now move to payback periods, contributed margin, working capital impact and how spend ladders into EBITDA" (Digiday, March 2026). In a retail media case, the payback period comes from the monthly view of cumulative contribution, and contributed margin is margin per dollar applied to ad revenue. Where brands prepay into wallets, the account balances their campaigns spend from, cash arrives before the campaigns deliver, which helps working capital. Contribution carries into earnings before interest, taxes, depreciation and amortization (EBITDA) after whatever shared costs finance allocates to the program.

Large retailers now describe retail media to investors in profit terms: profit for the Kroger Precision Marketing advertising business grew 24% in the second quarter, and Kroger CEO Greg Foran told investors it was the company's best retail media profit growth since 2021, adding that media monetization grew 88 basis points over the past year (Digiday, September 2026). Walmart's fourth-quarter fiscal 2026 release described "growing business mix contributions that continue to diversify profit sources, including advertising income and membership fees," and reported global advertising growth of 37% including VIZIO, with Walmart Connect in the U.S. up 41% (Walmart, February 2026).

The retailer's own pilot results give the board figures it can check. A pilot on onsite formats produces an actual margin per dollar to set against the plan, renewal rates for the first advertisers, and results brands can verify, such as incremental sales measured against a holdout group. Reporting those figures each quarter, together with M%G and contribution margin, keeps the program in front of the board as a business line from its first review.

Frequently asked questions

Does retail media revenue count as new money if brands move trade budgets into it?

The new money is the increase in a brand's total spend with the retailer. A brand that pays for retail media out of trade or promotion funds it already spent with the retailer adds to the retail media line and leaves its total funding flat. Compare each large brand's combined trade and media spend before and after launch, and count the difference.

Should the platform be paid a revenue share or a fixed license?

A revenue share keeps the platform cost in step with ad revenue, which protects the early months, and it takes a larger amount as the program grows. Under a fixed license, the fee is a large share of ad revenue in the early months and a falling share as ad revenue grows. Run the model both ways at the low, base and high revenue cases before choosing.

What ROAS should the business case assume for brands?

Use the ROAS the retailer's own campaigns show for each format and category, measured the way brands measure it. Before campaign history exists, start from published benchmarks for the same platform and format and replace them with the retailer's own results after the first quarter of campaigns.

Should shared costs be allocated to the retail media P&L?

Show the program with and without an allocation of shared costs. On incremental costs, the model answers whether to launch or scale. Fully loaded with a share of shared engineering, data and site costs, it lets the board compare retail media with the retailer's other business lines.

How should in-store screens enter the model?

Screens, players and installation are capital spend, so they enter the P&L as depreciation over the screens' useful life and the payback calculation as upfront cash. Model in-store revenue per store, per campaign or per day, since impression-based pricing needs in-store sensors to count who saw each ad.

Sources

1. Aswath Damodaran, "Margins by Sector (US)", NYU Stern School of Business, data as of January 2026. pages.stern.nyu.edu
2. Zacks Equity Research, "Target's Roundel Business Is Quietly Powering Profit Growth", Yahoo Finance, 8 July 2026. finance.yahoo.com
3. Osmos, "Unlocking homepage monetization: How India's leading online pharmacy achieved 20% ad revenue growth in one month". https://www.osmos.ai/success-stories/indias-leading-online-pharmacy-generates-484k-revenue-with-high-impact-display-formats-with-osmos (verified live 23 September 2026)
4. Mitchell Parton, "Retail media is critical to one of the nation's largest grocery chains", Digiday, 15 September 2026. digiday.com
5. Peter Wilmot (Shopsense), "Why Off-Site Retail Media Is Growing Twice as Fast as On-Site", Advertising Week, 26 February 2026, citing eMarketer data. advertisingweek.com
6. NielsenIQ, "How Brands Can Measure Incrementality and Prove Retail Media Impact", 24 July 2026. nielseniq.com
7. Krystal Scanlon, "The CFO moved upstream. Now agencies have to as well", Digiday, 4 March 2026. digiday.com
8. Walmart, "Walmart Releases Q4 FY26 Earnings", 19 February 2026. corporate.walmart.com