What is Commerce Media?

Commerce has always generated data about its customers. The question has been whether the business processing the transactions could profit from that knowledge.

Department stores figured it out first. Macy's and Sears, by the early 20th century, built a commercial model around exactly this: manufacturers paid for prime floor placement, end-cap displays, and catalog position because retailers controlled access to buyers that brands couldn't reach any other way. The retailer held the customer relationship. The brand paid for proximity to it. That was commerce media, before anyone had a name for it.

Loyalty programs in the 1980s and 90s made the data explicit. American Airlines launched AAdvantage in 1981; Kroger followed with its loyalty card in the late 1990s. Both built the first structured databases of individual purchase behavior at scale. Airlines sold miles to credit card companies and hotels. Grocery chains sold their transaction data to CPG brands that had never had direct visibility into how their products were bought or by whom. The transaction relationship had become a media asset. The infrastructure to activate it at speed didn't exist yet.

The internet changed that. Amazon's sponsored product listings, launched in 2012, applied the logic of the department store end-cap to a system with real-time targeting and closed-loop measurement. A brand could pay for placement, see exactly which clicks led to purchases, and prove return on ad spend against verified transactions. Retail media moved from backroom trade deals to a measurable, programmable channel.

Commerce media is advertising powered by first-party transaction data, and that model turned out to extend well beyond retail. Any business processing purchases at scale holds the same commercial advantage: verified data on what its customers buy, when, and how much they spend. It covers any touchpoint those businesses control: on-site search and category pages, checkout flows, post-purchase confirmation screens, in-store digital displays, connected TV, and off-site programmatic. The channel is not what defines it. The transaction relationship is.

That reach is reflected in the numbers. Commerce media grew to a $178.2 billion global advertising category in 2025, overtaking total TV advertising revenue for the first time, with 56% of advertisers planning to shift budgets into it over the next 12 months.

A retailer serving sponsored product listings to active shoppers is doing it. A bank surfacing card offers inside its mobile app is doing it. A grocery chain running targeted promotions on in-store digital screens is doing it. So is an airline embedding relevant deals into a booking confirmation email. Transaction data is the common thread.

How Commerce Media Works

The raw material is transaction data: what customers buy, when they buy it, how much they spend, and how they behave across the buying journey. This data is more predictive than demographic or interest-based targeting because it reflects actual behavior, not inferred preferences. A customer who just bought a last-minute flight to New York tells you more about their next likely purchase than their age, income bracket, or browsing history ever could.

A commerce media network makes that data actionable by connecting it to ad inventory across every touchpoint the business controls, on-site and off. Brands access that inventory and target audiences based on verified purchase history rather than third-party cookies or probabilistic audience IDs.

Because every impression exists inside a transaction environment, measurement works differently than in traditional display. Advertisers can see clicks, incremental purchases, verified customer acquisition cost, and return on ad spend calculated against real purchase events. The model closes the loop that search, social, and display leave open.

Commerce Media vs. Retail Media

Retail media is a subset of commerce media. Amazon Advertising, Walmart Connect, and Instacart Ads are retail media networks operating inside a single retailer’s ecosystem, drawing on that retailer’s first-party shopping data. Retail media was the category’s first chapter.

Commerce media is the broader category that applies the same data-driven, transaction-adjacent model to any business that processes purchases at scale. Banks, airlines, food delivery platforms, ticketing companies, telecoms, and rideshare apps now all operate commerce media businesses built on the same underlying logic.

The practical difference for advertisers comes down to reach and measurement fidelity. A retailer’s on-site inventory reaches people actively shopping that specific retailer. Commerce media extends targeting and closed-loop measurement across the full range of transaction-rich environments, including high-intent moments that no retail media network can access.

McKinsey projects the US commerce media market will exceed $100 billion by 2027. The median advertiser now works with six commerce media networks, up from four just a year ago, a sign of how rapidly the category has fragmented beyond the large retail walled gardens.

The verticals that have moved into commerce media

Commerce media’s expansion beyond retail accelerated sharply between 2024 and 2025, with four industries moving particularly fast.

Financial services built the earliest non-retail commerce media businesses. PayPal, Chase, Klarna, and Revolut have all launched advertising platforms built on purchase-level data. Klarna reached 100 million users across 724,000 merchants by Q1 2025 and built its ad business on top of that transaction network. PayPal draws on purchase history across its ecosystem, including Venmo, to offer audience targeting that no single-retailer walled garden can replicate.

Travel followed with some of the highest-profile launches. United Airlines debuted Kinective Media, marketing it as an airline industry first. Marriott International launched MARRIOTT MEDIA in mid-2025, backed by 237 million Bonvoy loyalty members and more than 200 targetable audience attributes. Pilot partners included PepsiCo, Visa, Uber, Starbucks, F1 THE MOVIE, Audible, American Express, and Resy (brands whose ads appear in an environment or platform where the advertised product or service is not directly related to the core content or audience’s primary intent), reaching travelers at the moments of highest purchase intent in the booking and travel journey.

Delivery and rideshare arrived earlier. Instacart generated over $1 billion in US advertising revenues in 2025. Uber, Lyft, and DoorDash each operate competing networks.

Ticketing and live events complete the picture. A customer buying concert tickets is in an active purchasing mindset, and their purchase data reveals genre preferences, income range, and geography in ways that general audience data can’t approach. Commerce media in this vertical was always a natural fit; the infrastructure to run it at scale arrived recently.

The unifying logic across all of these: any company with large-scale transaction data and owned customer touchpoints can run a commerce media business. The prerequisite is not being a retailer. It’s processing purchases. To execute this at scale, especially in regulated sectors like banking, commerce media utilizes data clean rooms (DCRs). This infrastructure allows advertisers to match their own audience segments with an operator’s transaction signals in a 'blind' environment, ensuring that Personally Identifiable Information (PII) is never decrypted or transferred, maintaining 100% GDPR and CCPA compliance.

The Transaction Moment™: Commerce media’s highest-performing window

Commerce media inventory spans the full customer journey. Sponsored product listings reach customers at the top of the funnel. Display units in checkout flows catch them mid-decision. The Transaction Moment is the specific window spanning product selection, cart or review, payment, and purchase confirmation where attention, intent, and trust peak. At this point, the customer is committed. The purchase decision is done. Attention is focused on the confirmation. 

Commerce media placements that activate this window consistently outperform standard formats. Transaction placements — spanning payments pages, cart and checkout upsells, and purchase confirmation screens — produce meaningful incremental revenue that sits outside the standard purchase flow. The reason is straightforward: the customer's decision is made, attention is focused, and the offer appears in a context of completed rather than uncertain intent. 

The performance difference against standard display is measurable. Transaction placements consistently deliver click-through rates an order of magnitude above standard display benchmarks. The conversion rate differential reflects what targeting on a confirmed, live transaction signal produces compared to reaching a general audience through interest-based or demographic segments.

At network scale, the performance advantage compounds. Placement decisions informed by cross-network transaction signals, aggregated across multiple operators without PII transfer, are materially more accurate than decisions made within a single operator's isolated dataset. The larger the transaction network, the better the signal quality for every placement decision across it.

How commerce media is measured

Attribution is the reason commerce media has pulled budget from display and social at the pace it has. The closed-loop model works because both sides of the transaction (the ad impression and the purchase event) exist within the same data environment.

ROAS (return on ad spend) is revenue generated per dollar of advertising investment. While standard retail media often relies on SKU-level attribution, broader commerce media enables ecosystem attribution. This tracks the 'halo effect': how a transaction in one category (e.g., booking a flight) drives a conversion in an unrelated category (e.g., signing up for a premium credit card). This is measured through hold-out testing, comparing the behavior of an 'exposed' group vs. a 'control' group to prove absolute incrementality. Because commerce media can verify purchases directly, ROAS figures are based on actual transactions rather than modeled or probabilistic attribution.

Incremental ROAS (iROAS) measures the lift in purchases attributable specifically to the ad, controlling for customers who would have converted anyway. This is the number that separates meaningful commerce media measurement from last-click attribution, which tends to credit conversions that would have happened without the ad.

Customer acquisition cost (CAC) tracks the cost of each verified new-to-brand buyer for advertisers using commerce media to expand their customer base rather than retarget existing ones.

Share of wallet matters most for non-endemic advertisers (brands that do not sell through the commerce platform running the ad). Here, measurement focuses on behavior change: did the customer increase spend with the brand, switch from a competitor, or take a specific action after seeing the ad?

The measurement advantage is why 56% of advertisers report planned budget increases into commerce media networks over the next 12 months, with most of those dollars coming from social and display budgets that cannot offer attribution at the same fidelity.

What commerce media means for each stakeholder

For brands, commerce media delivers qualified reach and measurable returns. Rather than targeting probabilistic segments, brands reach customers whose purchase history confirms they are relevant. A subscription fitness service advertising on a sporting goods retailer’s post-purchase confirmation page is reaching someone who just spent money on athletic gear. Transaction conversion rates reflect what reaching customers on a confirmed transaction signal produces at scale, consistently outperforming pre-purchase display and social placements.

For commerce media network operators (whether a retailer, airline, bank, or ticketing company), commerce media turns existing customer data and touchpoints into a revenue stream that did not exist before. Instacart generated over $1 billion in 2025 ad revenues. The best commerce media partnerships return the majority of value generated back to the operator, with full visibility into revenue, data, and performance metrics.

For customers, relevance is the value exchange. A customer's full transaction history tells a brand what a single purchase never could. Someone who books business-class flights several times a year and maintains premium subscriptions is a well-defined audience for a financial product or a luxury brand, whether they are completing a grocery order or booking a dentist appointment. The transaction profile is the signal. The current purchase is incidental. Commerce media’s first-party data foundation also operates without third-party cookies or shared PII, a structural privacy advantage over audience segments built on third-party tracking infrastructure that is disappearing.

The challenges worth understanding

Data quality and network depth. Commerce media network quality is determined by the quality of its transaction data. Networks with thin transaction history produce weaker targeting. Before committing to a network, advertisers should ask how many transactions it processes annually, how many years of purchase history it holds, and what its identity resolution methodology is.

Fragmentation across networks. With more than 200 commerce media networks now active in the US, managing campaigns and measurement across multiple platforms is operationally demanding. The category lacks the unified measurement taxonomy that paid search has had for two decades.

Non-endemic placement relevance. Non-endemic advertisers access strong intent signals but sometimes face relevance challenges. A luxury automotive brand advertising on a grocery delivery app may reach the right income demographic but miss on purchase context. The best networks solve this with behavioral modeling that goes beyond the immediate transaction category.

Confirmation page experience. Transaction inventory delivers strong performance precisely because customers are paying close attention at that moment. An aggressive or irrelevant offer on a confirmation screen can undermine purchase satisfaction. Networks that optimize for engagement quality over raw impression volume hold performance over time.

What to look for in a commerce media partner

The commerce media category has grown fast enough that more than 200 networks now compete for advertiser budgets in the US. Most offer variations on the same value proposition. The factors that actually differentiate performance:

• Transaction data at scale. A network processing billions of annual transactions produces categorically better audience signal than one processing millions. Scale also produces more reliable incrementality baselines.
• Real-time AI decisioning. Commerce media at the Transaction Moment runs on millisecond decisions. Networks that rely on rule-based targeting or batch-updated audience segments cannot operate at that speed. Networks that have operated at transaction scale for years build meaningfully better models than new entrants; look for documented tenure and transaction volume, not just technology claims.
• Measurement transparency. Incrementality testing, not just last-click attribution. A network that cannot isolate which conversions were incremental is likely quoting ROAS figures that include purchases that would have happened regardless.
• Partnership economics. The best partners offer full visibility into revenue performance and data, with economics structured so that operator revenue grows in proportion to network performance.
• Security and compliance posture. SOC 2 Type II, ISO 27001, GDPR and CCPA compliance are baseline requirements. Ask for certifications and audit reports, not just claims.

How Rokt fits in this landscape

Rokt built its commerce media capabilities specifically around the Transaction Moment,  the window spanning product selection through to purchase confirmation, where attention, intent, and trust peak. The Rokt Ecommerce Network operates across more than 33,000 active clients spanning retail, airlines, ticketing platforms, food delivery companies, and financial services, processing more than 10 billion transactions annually across 1.1 billion unique customers globally.

The network’s performance reflects what happens when targeting runs on confirmed transaction signals rather than browsing inference. Rokt Ads delivers a 4.03% click-through rate, 10 times the Google Display benchmark and 4 times Facebook. The 6.32% conversion rate on Rokt Ads placements reflects what post-transaction inventory produces when targeting is grounded in a live purchase signal rather than a probabilistic audience segment. For context, that CTR rivals high-intent search, despite running at the post-purchase stage rather than the pre-purchase discovery stage.

The AI that powers Rokt’s placement decisions has been trained on 1.95 trillion data points across the network. Every placement decision is informed by what customers who match this customer’s profile have done across the entire network, not just within one retailer’s isolated dataset. This cross-network signal quality is what separates performance at Rokt’s scale from what a single-operator commerce media build can produce.

For ecommerce partners, Rokt’s products, Rokt Catalog, Rokt Upcart, Rokt Pay+, and Rokt Thanks, monetize existing transaction touchpoints without requiring the operator to build any advertising infrastructure. Rokt returns $7 of every $8 of value generated back to the operator, with full visibility into revenue, data, and performance metrics.

For advertisers, Rokt Ads provides access to verified transaction audiences across all the verticals Rokt operates in simultaneously; a single buy that activates purchase signals from retail, travel, ticketing, delivery, and financial services at once. Rokt is trusted by more than 50% of the leading 200 ecommerce companies globally, including Live Nation, Macy’s, Fanatics, AMC Theatres, PayPal, Uber, Hulu, Staples, Albertsons, and HelloFresh, and maintains 99.992% uptime across 33,000+ active clients.

Frequently asked questions about commerce media

‍

What is the difference between commerce media and retail media?

Retail media refers specifically to advertising within retailer-owned ecosystems: Amazon Advertising, Walmart Connect, Target Roundel, Instacart Ads. Commerce media is the broader category: any advertising that uses transaction data and commerce touchpoints to reach customers, whether the operator is a retailer, bank, airline, food delivery platform, or ticketing company. All retail media is commerce media. The reverse is not true.

What types of companies can run a commerce media business?

Any company that processes transactions at scale and owns direct customer touchpoints. Retailers were first to build formal commerce media networks, but financial services companies, airlines, food delivery platforms, ticketing companies, telecoms, and rideshare apps have all launched their own. The prerequisite is transaction data, not a product catalog.

What is a non-endemic advertiser in commerce media?

A non-endemic advertiser is a brand that does not sell products through the commerce platform it is advertising on. The success of non-endemic advertising in commerce media is driven by signal proximity. A travel rewards card on a grocery app works because the data signal (high grocery spend) is a proxy for household income and stability: a 'financial lifestyle' signal that is more predictive of creditworthiness than a simple search for 'best credit cards.' Non-endemic advertising works because transaction data reveals purchasing behavior and lifestyle patterns that are valuable to many brands beyond the immediate product category.

How is commerce media measured?

The standard metrics are ROAS based on verified transactions, incremental ROAS (the lift specifically attributable to the ad after controlling for organic conversions), and customer acquisition cost for new-to-brand buyers. Commerce media measurement advantage over social and display is that purchase verification happens within the network itself.

What is first-party data and why does it matter for commerce media?

First-party data is information a company collects directly from its own customers: purchase history, transaction records, in-app behavior. Commerce media networks built on first-party data reach customers without third-party cookies or shared personal information. Purchase history is more predictive of future purchases than inferred interest signals.

What is the Transaction Moment in commerce media?

The Transaction Moment is the specific window spanning product selection, cart or review, payment, and purchase confirmation where attention, intent, and trust peak. The purchase decision has already been made. Commerce media placements in the Transaction Moment consistently outperform pre-purchase placements.

What pricing models does commerce media use?

Commerce media networks use cost-per-click (CPC), cost-per-acquisition (CPA), and revenue share arrangements. Post-transaction placements that drive subscriptions typically run on CPA. Sponsored product placements within shopping flows typically run on CPC.

What is the difference between on-site, off-site, and in-store commerce media?

On-site commerce media runs within the commerce platform itself: sponsored listings, checkout flow display units, post-transaction confirmation pages. Off-site uses the platform first-party audience data to reach customers across external channels. In-store commerce media includes digital screens and receipt placements.

Why is commerce media growing faster than other digital advertising categories?

Two factors drive disproportionate growth. Attribution fidelity: commerce media closes the loop on conversions using verified purchase events rather than modeled attribution. Data durability: commerce media first-party transaction data is unaffected by cookie deprecation or IDFA restrictions.

How does the Transaction Moment compare to standard display and paid search?

Standard display targets audiences by demographics or inferred interests; paid search captures declared intent at the keyword level. The Transaction Moment operates on a different signal entirely: confirmed purchase action. A customer who has just completed a transaction has demonstrated willingness to spend in a specific context, providing a more reliable signal than either browsing behaviour or search queries.

The practical outcome is measurable. Post-transaction placements consistently produce click-through rates an order of magnitude above standard display benchmarks. Conversion rates reflect the difference between reaching someone who might be interested and reaching someone who has just confirmed active purchasing intent.

How does commerce media handle 'cookie deprecation'?

Commerce media is functionally 'cookie-less' because it relies on deterministic first-party data. Unlike social or display ads that use probabilistic 'lookalike' models based on browsing history, commerce media uses confirmed transaction events tied to a persistent internal ID (like a hashed email or loyalty number). This makes the data durable against privacy changes in Chrome or iOS.

What is the difference between 'endemic' and 'non-endemic' in this context?

Endemic ads promote products sold directly on the host platform (e.g., a toothpaste ad on a grocery site). Non-endemic ads promote products or services that exist outside the platform (e.g., an insurance ad on a car-buying site). Commerce media’s unique advantage is its ability to use 'transaction signals' to make non-endemic ads feel contextually relevant to the user’s life stage.