← Back to homeKnowledge base

    Commerce media glossary

    The concepts you need to understand to monetize your digital inventory. Clear definitions, why they matter, and how they connect to the platform.

    Commerce media fundamentals

    The concepts that define the category: what commerce media is, how it differs from retail media, and why it matters for your business.

    Commerce media

    Key term

    Commerce media is the monetization of your own advertising inventory —on and off your platform— leveraging the first-party data and purchase intent you already have. It's not an e-commerce module: it's a business unit with its own P&L, its KPIs and its team. Global leaders like Mercado Ads and Amazon Ads monetize between 5% and 8% of their revenue in media; most commerce in LatAm, less than 0.5%. That gap is the business. Commerce media spans retailers, marketplaces, banks, telcos, super apps and airlines: anyone with their own audience, data and purchase intent can have a media business.

    Retail media

    Key term

    Retail media is the origin of the category: the monetization of a retailer or marketplace's advertising inventory, where transactional purchase data is the signal that decides which ad is shown. It was born in retailers like Amazon, Walmart and Mercado Libre, and that's why the name carries 'retail'. Today the category evolved into commerce media: the same engine runs in banks, telcos and super apps, not just stores. Retail media remains the most mature case and the benchmark to size the opportunity, but it's no longer the only vertical where the business exists.

    Related to:Verticals

    Retail media vs. commerce media

    Key term

    The distinction is about scope, not technology. Retail media refers to the media business built on a retailer or marketplace's inventory and data. Commerce media is the expanded category: it includes retail media but also banks, telcos, super apps, airlines and any platform with its own audience and purchase intent. The engine is the same —ad serving, decisioning models, incrementality measurement—; what changes is the inventory and the brands that buy it. Saying commerce media instead of retail media reflects that the best media business in the region isn't always in a store: in LatAm, banks and telcos hold intent data more insightful than retail.

    Related to:Verticals

    GMV (Gross Merchandise Volume)

    Key term

    GMV is the total value of merchandise sold through a platform in a period. It's the base metric to size the commerce media opportunity: if a retailer books US$ 200M in digital GMV per year and monetizes at 0.5%, the media line is US$ 1M (nearly invisible). If it monetizes at 5%, that's US$ 10M; at 8%, US$ 16M. GMV isn't the retailer's net revenue (it includes commissions, returns, etc.), but it's the most used benchmark to calculate how much advertising inventory exists and how much can be captured.

    Related to:Calculator

    First-party data

    First-party data is the information a platform collects directly from its own users: what they bought, what they searched for, where, when, how often. It's the asset that makes commerce media inventory valuable, because it lets you decide which ad to show each person based on their real purchase intent. Unlike third-party data (bought or shared), first-party data is exclusive, precise and increasingly valuable in a cookieless world.

    Ad formats

    The formats that make up commerce media inventory and how they differ in placement, intent and objective.

    Display Ads

    Display Ads are graphical ads (banners, rich media) that occupy dedicated advertising spaces within the platform: the home page, category pages, the cart, the product page. Unlike Sponsored Products, they don't integrate with the organic listing: they live in dedicated spaces that previously showed nothing or editorial content. That's why they're pure incremental revenue: every impression sold is a space that previously generated no money.

    Related to:Onsite

    Channels: onsite, offsite and in-store

    Where advertising inventory lives and how it extends beyond your own platform.

    Onsite

    Key term

    Onsite is the advertising inventory that lives inside your own platform: Sponsored Products in search results, Display Ads on the home page, Sponsored Brands on category pages. It's the highest-intent channel and the starting point of any commerce media business, because the user is already on your digital property with purchase intent. Onsite is where you prove the inventory converts and where the first brands renew.

    Related to:Onsite

    Offsite

    Offsite is the advertising inventory that extends outside your platform, using your first-party data to reach your users on other sites and apps. It lets brands keep pursuing a shopper who already visited you but didn't convert, or reactivate inactive customers. Offsite expands the available inventory beyond your own pages and captures budget that would otherwise go to social media or generic display, but with the advantage of the intent signal only you have.

    Related to:Offsite

    In-store

    In-store is the monetization of advertising inventory within the physical point of sale: digital screens, product displays, audio, digital receipts. It connects the digital and physical worlds: it uses online first-party data to personalize what's shown in the store, and measures impact with the same incrementality logic. For retailers with an omnichannel presence, in-store turns physical traffic —previously unmonetized— into an extension of the commerce media business.

    Related to:In-store

    Measurement and metrics

    The metrics that separate a report that looks good from one that renews the contract.

    Incrementality

    Key term

    Incrementality measures the sales that wouldn't have occurred without the ad, using a control group: a set of users comparable to the one that saw the ad, but that didn't see it. The difference in sales between the two groups is the incremental revenue. It differs from traditional attribution (last-click), which counts all sales that came through the ad —including those that would have happened anyway. A 16x ROAS includes sales that were going to happen regardless; a 3x incremental is lower but defensible: if you can show how you measured it, the advertiser renews. Incrementality is the central metric of the report at AndesML, not a premium add-on.

    ROAS (Return on Ad Spend)

    ROAS is the return on advertising spend: how many dollars of sale are generated for every dollar invested in advertising. A 16x ROAS means that for every US$1 in ads, US$16 in sales are attributed. The problem is that traditional attribution counts sales that would have occurred anyway, with or without the ad, inflating the number. That's why ROAS looks good on a dashboard but doesn't answer the question that matters: how much did I grow thanks to this investment? For that, incrementality is used.

    CTR (Click-Through Rate)

    CTR is the percentage of users who click on an ad out of the total who see it. It's a basic performance metric that indicates how relevant the ad is to the audience, but it doesn't measure business outcome: a high CTR doesn't guarantee sales. In commerce media, CTR is used as an optimization signal (which ads resonate) but not as a final success metric —for that, incremental sales and incremental ROAS are used.

    Attribution

    Attribution is the method that assigns credit for a sale to one or more advertising touchpoints. Last-click attribution —the most common— gives all the credit to the last ad before the purchase. It's simple but overestimates impact, because it counts sales that would have happened anyway. Incrementality is a more rigorous alternative: instead of attributing, it isolates the causal effect of the ad with a control group.

    Operation and monetization

    How a commerce media business is built and operated: the team, the playbook and the decisions that move the number.

    Ad server

    The ad server is the technical piece that decides which ad is shown on each impression, delivers it and records the outcome. It's necessary but not sufficient: the ad server is the tool, not the business. Many retailers buy an ad server thinking that's enough, and discover they lack the team to operate it, the playbook to teach how to sell the inventory, and the decisioning models to optimize the result. An ad server without a team is a depreciating asset.

    Related to:Platform

    Media kit

    The media kit is the commercial material your sales team uses to sell inventory to brands: deck for advertisers, rate card, available formats, measurement methodology and case studies. In commerce media, the media kit competes against those of Mercado Ads or Amazon Ads, so it must measure up. A well-built media kit is what lets your commercial team go out to sell with arguments and not empty-handed.

    Self-service

    Self-service is the ability for brands to buy advertising inventory on their own, through a portal with card payment, without sales team intervention. It serves the long tail of small and medium advertisers that don't justify a dedicated seller but that, together, represent a significant portion of revenue. Self-service scales the business without scaling the commercial team.

    Related to:Platform

    Monetization

    Monetization, in commerce media, is the process of turning traffic and first-party data into advertising revenue. It's measured as the percentage of GMV (or revenue) captured in the media line. The global benchmark is 5–8%; LatAm today is under 0.5%. Monetizing isn't just placing ads: it's packaging formats, pricing them, selling them to brands, measuring incrementality and repeating it every month. That's why it's a new business, not a module.

    Related to:Calculator

    You don't have to know all this. We already do.

    Commerce media has a lot of moving parts and new jargon. The good news is you don't have to figure it out alone: our team has already built and operates this business inside the largest retailers in the region. Let's talk and we'll accelerate yours.

    The diagnosis takes 45 minutes and you leave with an estimated revenue range for your business, whether or not you use AndesML.