Retail Media Audience Monetization: How Retailers Build, Package and Price Audience Products

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

Retail media audience monetization is how a retailer earns ad revenue by offering its shoppers to brands as an audience they can target, on its own pages and beyond them. The retailer groups shoppers into segments built from purchase and browsing behavior, then offers those segments as targeting on its own site and app, on offsite channels such as Meta, Google and TikTok, and in clean-room collaborations with brands, each priced for what it delivers.

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.

Every audience product sits on consented first-party data that the retailer has already collected and enriched as a targeting layer for ad revenue. The product itself has five parts: a shopper segment a brand can buy, the package it comes in, its price, the channels it runs on and the closed-loop sales reporting that comes back.

What is retail media audience monetization?

Retail media audience monetization is the practice of turning a retailer's shopper segments into ad products that brands pay to reach, on the retailer's own inventory and beyond it. Osmos supports both halves of that model: audience-targeted placements through its onsite retail media apps, and audience extension to Google, Meta and TikTok through its offsite retail media apps, with the retailer setting the pricing in each case.

Audience product is a defined, repeatable shopper segment (for example, "bought baby formula in the last 60 days") that a brand can switch on as targeting, with agreed rules for where it runs, how it is priced and how results are reported.

Product Ads, the first layer of retail media, are priced on the shelf position, so a brand pays for a place in search results or on a category page when a shopper is there, and the retailer earns when traffic arrives and space is free. With an audience product, the brand pays to reach a defined group of shoppers wherever the retailer can serve it, so the product keeps earning when the pages are full or the shopper is browsing somewhere else. Retailers treat audience products as the next layer of retail media monetisation once their onsite placements are well filled.

Retail data monetization and "retail access monetization" describe the same model. The retailer's revenue comes from media placements and audience products that reach its shoppers, with first-party data as the targeting layer that decides who sees what.

How are retailers monetizing customer relationships without ads?

Retailers monetize customer relationships without running an ad on their own pages by letting brands reach their purchase-based audiences on other media, by collaborating with brands on audience insight and measurement inside data clean rooms, and by offering managed outcome programs to advertisers who would never run campaigns themselves. Each of these routes still ends in paid media or a paid service, and each works without adding a slot to the retailer's site.

Offsite audience extension. A brand reaches a retailer's "category buyers" segment on Meta, Google, TikTok or programmatic channels, and Osmos runs the campaign on the retailer's own channel seats, so the retailer earns on the media and keeps the relationship and the margin.

Clean-room collaboration. A brand matches its own customer list against the retailer's shoppers to size the overlap, find gaps or plan a campaign, without either side exchanging row-level records. The brand pays for the overlap study and for the campaign it then runs on the matched segment.

Managed outcome programs. Smaller advertisers, such as regional brands or local services, pay for a fixed outcome. Osmos runs these as Guaranteed Outcomes: flat-fee packages for a guaranteed click or visit count, delivered across Meta, Google and DV360. The margin goes to the retailer, and its side needs no extra ad-operations headcount.

Loyalty programs and memberships feed every one of these routes, because the logged-in, repeat-purchase data they create is what makes a retailer's segments precise enough for a brand to pay for.

Which audience products can a retailer build?

A retailer can build retail advertising audiences from any behavior it records with consent: what shoppers buy, how often, what they search for and browse, which brands they switch between, and where they shop. The most useful retail audiences rest on actual transactions, which a brand cannot reproduce from its own data or from a social platform's interest categories.

Costco is a public example of a retailer naming its audience products. PPC Land reported in April 2025 that Costco offers Yahoo DSP (demand-side platform) buyers three segments, "Frequent Shoppers, Tenured Shoppers, and Category Shoppers", built on "extensive first-party data from its 80+ million US cardholders" (PPC Land, April 2025).

How each signal is collected and scored is covered in the guides to contextual and behavioral targeting and AI-powered ad serving and behavioral targeting. Retailers with thin data in some categories can enrich first-party data before turning it into segments. In-store screens and shelves are targeted by store, day and hour, a model covered in in-store retail media monetization.

How should a retailer package audience products?

Brands and agencies plan faster from a short, named catalog of standard segments plus a custom option, and the catalog spares the retailer's ad team from rebuilding the same segment for every campaign.

1. Standard versus custom. Standard segments (category buyers, lapsed buyers, in-market shoppers) cover most briefs and can be priced from a rate card. Custom segments, built to a brand's own definition, take analyst time and carry a higher price or a minimum spend.
2. Recency and size tiers. "Bought in the last 30 days" is a smaller, warmer audience than "bought in the last 12 months." Offering both, priced differently, lets a brand choose between precision and reach.
3. Channel scope. Each product states whether it runs onsite only, offsite only or both. A segment offered on both lets the retailer show one result across channels.
4. Bundles. Audience products sit well inside multi-format packages: an onsite display flight, an offsite social flight and Product Ads on the same segment, booked as one plan. The guide to retail media ad packages covers how to assemble and price those bundles.
5. Approval and suppression rules. The retailer decides which categories and advertisers can use which segments, and which shoppers are excluded (recent purchasers of the advertised item, opted-out shoppers, sensitive categories).

What are the ideas for monetizing onsite traffic with retail media ads?

Retailers monetize onsite traffic by running Product Ads on search and category pages and by adding audience targeting to the display and video placements they already run. Osmos's onsite retail media runs Product Ads, display, video, non-endemic and custom formats from one ad server, with CPC (cost per click), CPM (cost per thousand impressions), CPD (cost per day) or fixed-tenancy buying across inventory and formats.

Product Ads on high-intent pages. Search results, category pages and product pages carry the most purchase intent. Product Ads are assembled from the catalog listing, so the retailer's levers are placement, relevance and bid floors, and revenue grows by putting ads on more queries and pages, raising floors to the value of each placement, adding audience targeting to keywords and bringing in more advertisers. Osmos ships relevance ranking and retailer-set bid floors with its Product Ads engine.

Audience targeting on display and video. A homepage banner shown to everyone earns a run-of-site price. The same banner shown only to a brand's lapsed buyers earns more per impression, because the brand is paying for the segment as well as the space, and that is how an audience product raises the value of inventory the retailer already runs.

Non-endemic advertisers. Banks, telcos and travel brands want a retailer's audiences even though they are not on the shelf. Osmos's non-endemic ads use the same workflow as endemic campaigns, and retailers choose which surfaces accept them, typically checkout, coupon and post-order screens.

Custom formats for awareness budgets. A format built to match the retailer's own design, such as a category-page takeover or a homepage video unit, draws brands' awareness budgets, and Osmos's custom ad formats inherit the retailer's design system and go live within a couple of hours of a spec brief, with build time varying by complexity.

Tail and torso advertisers. Osmos gives the many suppliers who have no time to run campaigns self-serve campaign tools under the retailer's own brand, so the retailer's team can bring in more advertisers than it could manage by hand; for a marketplace, the third-party sellers are the largest pool of them.

How does offsite audience extension turn a segment into revenue?

Offsite audience extension turns a segment into revenue by delivering a retailer's audience to channels such as Meta, Google and TikTok, where the brand pays for media bought on that audience and the retailer earns its margin on every campaign, then proves results against its own sales data. It is the route that lets a retailer reach its shoppers when they are not on its site, which is also how a brand finds new retail audiences for its products.

Osmos's offsite retail media runs these campaigns on the retailer's own channel seats, the retailer sets the markup built into the CPM, and Google, Meta and TikTok are live today, plus DV360 and PubMatic for programmatic buying. Segments are built without code, identity-resolved and consent-checked against GDPR (General Data Protection Regulation) and CCPA/CPRA (California's privacy laws) before use, and the data stays in the retailer's environment. Conversions are deduplicated at SKU (stock-keeping unit) level across channels, with attribution windows the retailer can set per channel, such as 7, 14 or 30 days after a click.

Large retailers show the same pattern publicly. Walmart Connect reports a 28% average increase in online attributed sales among connected TV (CTV) audiences, compared with the same audience before exposure, across 133 offsite CTV campaigns run between February 2024 and January 2025, measured over a 14-day window (Walmart Connect). Because the measurement rests on purchase data, a brand can see sales movement in the exact audience it paid to reach. It is Walmart's own, self-reported pre/post comparison, so it shows attributed sales; a measure of incrementality would need a holdout test.

Kroger took a service-led route. In September 2025, Kroger Precision Marketing launched "a suite of new managed-service offsite programmatic capabilities", noting that "more brands are applying purchase-based audiences across streaming media" (The Shelby Report, September 2025). The same report says the service "simplifies campaign execution for smaller marketing teams", and a retailer that runs the campaign for those teams captures budget that would otherwise stay unspent.

India's leading baby and kids retailer shows the onsite-to-offsite path in one engagement with Osmos. Its ad revenue stood at Rs 45 Lakhs in January 2025. By June 2025 it had reached ₹1.3 Crores a month, a 190% increase in six months, with product listing ad (PLA) revenue up 30% and display ad revenue up 47% month on month in June 2025. With that onsite base in place, the retailer went offsite on Meta with Osmos from May 2026 and earned ₹8 lakh across 5 advertisers in its first three months of offsite.

New-to-brand sales measure trial of the advertiser's brand. A weekly shopper who has never bought a given brand counts as new-to-brand for that advertiser while already being a loyal customer of the retailer. Choosing which channels to open first is covered in which offsite channels to offer first, and the operational changes offsite brings are covered in what breaks when you go offsite.

When does an audience product need a data clean room?

An audience product needs a data clean room when two parties must combine their own records to create it, such as matching a brand's customer list against the retailer's shoppers, and much less often for measurement the retailer can run inside its own environment. The IAB Tech Lab describes a clean room as "a secure collaboration environment which allows two or more parties to leverage data assets" without exposing the underlying records to each other, as quoted by PPC Land.

Data clean room is a controlled environment where a retailer and a brand can match and analyze their first-party data together, with rules that stop either side from seeing or extracting the other's row-level records.

The IAB makes the same point in its September 2026 guidance: "Data clean rooms are increasingly important for privacy-enhanced collaboration across the commerce ecosystem, but they aren't the right solution for every data or measurement need," it states (IAB, 29 September 2026). Its framework, as reported by PPC Land on 9 October 2026, sorts 14 common commerce use cases into two groups (PPC Land, October 2026):

• Five enabled by a clean room: audience overlap analysis, data enrichment, bring-your-own-data (BYOD) targeting, suppression and exclusion, and retailer audience activation.
• Nine enhanced by a clean room: including attribution, incrementality, media mix analysis, lookalike modeling, audience discovery and customer segmentation, which "can often be performed using existing measurement, analytics, or reporting tools."

Closed-loop attribution and incrementality reads on campaigns that run inside the retailer's own environment sit among the nine that often work with existing tools. A clean room earns its cost when a brand wants to bring its own customer data into the targeting, measure overlap before a large commitment, or suppress its existing customers from an acquisition campaign. Retailers that need one for these brand-data use cases typically pair their retail media platform with a dedicated clean-room provider. Questions of who owns which data in those partnerships are covered in the future of data ownership in retail media.

How do retailers price audience products?

Retailers price audience products with the same core models they use for other media, CPM for audience-targeted impressions, CPC for Product Ads, and fixed fees for packages and custom work, adjusted for segment precision, recency and channel. Audience rates are negotiated between each retailer and its brands, and no industry benchmark publishes them.

Audience pricing holds up in negotiation when a small, recent, high-intent segment costs more per impression than a broad one, and when the brand sees closed-loop sales results for the segment it paid the premium for. Offsite media carries pass-through costs, so offsite prices should be set for the margin they leave the retailer; the case for judging retail media on margin is set out in the retail media business case. Osmos leaves pricing with the retailer on both sides: onsite buying models and floors are set by the retailer, and offsite campaigns carry the markup the retailer chooses.

Why do advertisers pay for retailer audiences?

Advertisers pay for retailer audiences because they are built on what shoppers actually buy, and because the retailer can show the sales that followed an ad in the same data.

Accuracy from transactions. A segment of people who bought a product is more precise than a segment modeled from interests or browsing elsewhere. Brands pay for that precision, and it is the retailer's main advantage over open-web targeting.

Proof against sales. Closed-loop reporting links an impression to a purchase in the retailer's own records. Brands renew their budgets when the retailer shows closed-loop sales results for the segment they bought.

Steadier revenue across the year. Audience products earn whenever a brand wants to reach the segment, including on offsite channels in quiet weeks. Product Ads rise and fall with site traffic, so audience revenue smooths the retailer's ad income across the year.

Reach beyond the page. Audiences can be activated across many channels, so ad revenue can grow past the fixed number of ad positions on the retailer's pages while the shopping experience stays uncluttered.

Privacy and compliance by design. Retail media platforms activate segments without exposing personally identifiable information (PII), with consent checks before a segment is used. Brands increasingly ask for this before they commit budget, and a retailer that can show it shortens the approval cycle.

What does a retailer need in place to run audience products?

A retailer needs consented first-party data with clear governance, identity resolution that connects shoppers across devices and channels, a segment builder its team can use without engineers, closed-loop measurement, and enough operating capacity to serve advertisers at scale. Osmos covers the activation and measurement side through modular apps that sit over the retailer's existing stack, so the retailer can add audience products without replacing its ad server or data platform.

Consent and governance. Every segment needs a lawful basis, an approval owner and a list of categories it cannot be used for. Consent checks should run before a segment is activated.

Identity resolution. Matching logged-in shoppers, app users and hashed emails to one profile decides how much of a segment can be reached offsite. Retailers that operate several store brands can also pool audiences across them; the guide to running a multi-banner retail media network covers how.

A segment builder and a catalog. The ad team needs to update segments on a schedule to keep the standard catalog current as shopping patterns change.

Closed-loop measurement. Reporting has to connect onsite and offsite delivery to orders at SKU level, with attribution windows agreed with the brand before the campaign starts.

Operating capacity. Self-serve tools, managed programs or both let the retailer serve more advertisers than its ad team could handle alone.

On the Osmos platform, onsite formats run on Osmos's own ad server while offsite campaigns reach DV360 and PubMatic alongside Google, Meta and TikTok, all reported against the same sales data.

Frequently asked questions

How do you launch Product Ads on an ecommerce site?

Launch Product Ads (sponsored listings) on the pages that carry the most intent, usually search results and the top category pages, set eligibility rules and bid floors, and invite brands already in the catalog to fund ads on their own listings. Because Product Ads are assembled from existing product data, there is no creative to produce, so the launch work is integration, policy and advertiser onboarding. Osmos runs Product Ads on its own ad server on the retailer's domain, so the retailer does not have to replace its existing site stack.

How do marketplaces earn ad revenue from third-party sellers?

Marketplaces earn ad revenue from third-party sellers by letting them fund Product Ads and display placements on their own listings, usually from a prepaid ad wallet, while the marketplace sets floors, eligibility and relevance rules. Sellers are the largest and most fragmented advertiser group, so self-serve tools and simple packages bring them in. The build decisions behind a marketplace ad business are covered in how marketplaces build their own ad platform.

How large should a segment be before a retailer offers it?

A segment should be large enough to deliver a campaign's planned impressions over its flight at a sensible frequency, which depends on the channel. Onsite, a small segment of frequent visitors can work. Offsite, retailers check the matched size on the destination platform before confirming a booking, because only part of a segment matches to a platform's users.

How do commerce media networks generate revenue for businesses?

Commerce media networks generate revenue by offering ad placements and audiences built from transaction data, the same model as retail media but run by businesses beyond traditional retailers, such as delivery apps, travel sites and payment providers. The guide to commerce media explains how those networks differ from retail media networks and where they overlap.

Turn your shoppers into audience products

Audience products add a second layer of retail media revenue on top of Product Ads, and it keeps earning when pages are full and shoppers are elsewhere, provided the retailer can package and price each product and report the sales it drove. To see how Osmos's onsite and offsite apps run audience products on a retailer's own data and inventory, book a demo with Osmos.

Sources

1. IAB, "Data Clean Rooms in Commerce: Common Use Cases", 29 September 2026. iab.com
2. PPC Land, "IAB says 9 of 14 commerce use cases often work without a clean room", 9 October 2026. ppc.land
3. Walmart Connect, "Offsite media", data period 1 February 2024 to 31 January 2025, accessed 12 October 2026. walmartconnect.com
4. The Shelby Report, "Kroger Precision Marketing expands managed service capabilities", 23 September 2025. theshelbyreport.com
5. PPC Land, "Yahoo and Costco partnership leverages first-party data for advertisers", 21 April 2025. ppc.land