Retail media is an advertising mechanism in which retailers provide their digital touchpoints and in-store touchpoints as advertising surfaces, selling them to brands and manufacturers. It encompasses diverse formats including search ads, sponsored products, display, video, and in-store signage. As privacy regulations and tracking restrictions intensify, retail media is rapidly gaining prominence as a high-precision targeting channel built on retailers' first-party data.
Retail Media refers to an advertising mechanism in which retailers offer their digital touchpoints (e-commerce sites, apps) and in-store touchpoints (in-store digital signage) as advertising surfaces, selling them to brands and manufacturers. It encompasses diverse formats—search ads, sponsored products, display ads, video ads, and in-store digital signage—and in addition to on-site delivery, off-site delivery leveraging the retailer's first-party data across external media has also become a standard practice. In recent years, the broader concept of "Commerce Media" has emerged, encompassing retail media as its primary subcategory.
The backdrop is the tightening of privacy regulations, tracking restrictions imposed by operating systems and browsers, and the growing instability of ad measurement and targeting that depends on third-party cookies. As the signals that traditional digital advertising relied upon erode, the value of first-party data (customer data collected directly by the retailer) held by retail companies has surged. Information such as "what consumers purchased," "which product pages they browsed," and "at what time of day they tend to buy" constitutes high-precision signals that directly reflect purchase intent—fundamentally different in quality from generic demographic data.
The advancement of Generative AI is also providing tailwinds. Applications in the creative domain are expanding, including automated generation of ad creatives and improved efficiency in producing personalized ad copy. Meanwhile, integration with Demand Forecasting AI and Dynamic Pricing is more accurately characterized as an advanced use case of retail optimization rather than as part of the retail media definition itself.
Based on industry standards from the IAB and similar bodies, retail media is broadly classified into three formats:
CDP (Customer Data Platform), data clean rooms, and measurement infrastructure are increasingly being adopted as integrated data foundations to manage these components.
For advertisers—primarily manufacturers and brands—the greatest benefit is the targeting precision backed by purchase data. Delivery at granular levels, such as "users who purchased a competing product within the past 30 days" or "users who regularly subscribe to a specific category," becomes possible. Closed-loop measurement enables direct measurement of sales contribution and ROAS as KPIs. This is also an easy-to-evaluate model from an AI ROI perspective.
At the same time, there are points to be mindful of. Because data formats and measurement methodologies differ across retailers, cross-retailer performance comparisons tend to be difficult. There is a tendency for data to become siloed within each company's proprietary walled gardens, and measuring incrementality remains an industry-wide challenge.
From a privacy protection standpoint, compliance with data protection legislation across jurisdictions is indispensable—including GDPR (EU), CCPA/CPRA (California, US), and PDPA (Thailand). Specifically, consent management, opt-in/opt-out compliance, data governance frameworks, data collaboration through clean rooms, and proper access controls are required.
This advertising model, centered on purchase data as a "signal of intent," is driving a structural shift that tilts the balance of power in digital marketing toward retailers. Beyond mere diversification of advertising revenue, it holds the potential to redefine the competitive advantages of retail itself—through collaborative marketing with brands, integration with shelf space optimization, and more.



A2A (Agent-to-Agent Protocol) is a communication protocol that enables different AI agents to perform capability discovery, task delegation, and state synchronization, published by Google in April 2025.

Acceptance testing is a testing method that verifies whether developed features meet business requirements and user stories, from the perspective of the product owner and stakeholders.

AES-256 is the highest-strength encryption algorithm using a 256-bit key length within AES (Advanced Encryption Standard), a symmetric-key cryptographic scheme standardized by the National Institute of Standards and Technology (NIST).

A mechanism that controls task distribution, state management, and coordination flows among multiple AI agents.

Agent Skills are reusable instruction sets defined to enable AI agents to perform specific tasks or areas of expertise, functioning as modular units that extend the capabilities of an agent.