United States retail media advertising reached approximately $55 billion in 2024. Three examples are: Amazon generated $56.22 billion in global advertising services revenue for the same period, Walmart Connect that generated $4.4 billion, and Target Roundel generated $649 million. Every dollar of that revenue came from advertisers using and refining behavioral profiles assembled from personal digital information. Consumers whose digital participation made those profiles possible received none of it. This paper identifies where the extraction is realized.
Papers One through Four documented extraction from participation across four domains: platform attention and behavioral data, gaming and creator economies, compelled educational participation, and the continuous ambient monitoring of domestic life. Each paper established the same structural condition: participation originates in persons, generates commercially valuable data, and is captured by platforms that don't recognize the person as the origin of what they contribute. Paper Five asks where that data goes.
The answer isn't that it sits in platform databases generating abstract value. The behavioral profiles assembled from years of digital participation have a specific commercial destination: the moment of purchase. Every search query, every social media interaction, every connected television viewing session, every voice assistant command, every loyalty program transaction contributes to a behavioral model of commercial intent that is applied (in real time, at the point of sale) to shape the commercial environment through which a consumer reaches a transaction.
This isn't a speculative claim about what platforms might do with behavioral data. It's a description of a documented, disclosed, actively marketed commercial architecture that generated approximately $55 billion in U.S. advertising revenue in 2024. Retailers and platform operators sell access to the purchase moment (the ability to determine what appears in a consumer's search results, what sponsored products appear before organic listings, what deals are presented, and what urgency signals are deployed) to advertisers who pay for that access, because the behavioral profiles that make it possible are commercially valuable. The consumer who generated those profiles is the last party the system is designed to benefit.
Amazon: The Dominant Purchaser
Amazon is the largest retail media network in the United States by a margin that makes every other network secondary. Amazon Advertising generated $56.22 billion in global advertising services revenue for fiscal year 2024, up 20% year over year, with Amazon representing approximately 77% of all U.S. digital retail media advertising spending. The fourth quarter of 2024 alone generated $17.3 billion, the first quarter in which Amazon's advertising business exceeded $17 billion in a single period.
Amazon's advertising business operates at the intersection of two things its platform uniquely possesses: the world's largest product catalog and the world's most comprehensive behavioral record of consumer purchase intent. This isn't shelf placement or traditional merchandising. This is individualized interception based on accumulated behavioral history. A person who searches for a product on Amazon has already expressed commercial intent more precisely than any prior advertising medium could capture. They're not browsing a social media feed where a product might be relevant, they're searching for a specific thing, in a specific moment, with a demonstrated intention to purchase. That intent signal (assembled from years of search history, purchase history, browsing behavior, cart additions and abandonments, and connected device interactions) is the input to an advertising system that charges brands to appear in front of that person at the exact moment their intent is highest.
The Federal Trade Commission (FTC) documented on the public record what this architecture produces in practice. The FTC's antitrust complaint against Amazon, filed in September 2023, and proceeding after Amazon's motion to dismiss was denied in October 2024, alleges that Amazon degrades the customer experience by replacing relevant organic search results with paid advertisements that worsen search quality, biases its search results to preference Amazon's own products, and uses anti-discounting measures that prevent sellers from offering lower prices elsewhere, thus keeping prices higher across the internet. The FTC further alleges that Amazon uses a secret pricing algorithm called Project Nessie, described in the complaint as a tool for identifying when Amazon could raise prices without losing sales. These aren't allegations of incidental misconduct; they're regulatory findings that the commercial architecture of the world's largest retail platform is organized to extract maximum value from consumers at the point of purchase. It's using behavioral data assembled from those consumers' participation to determine what they see, what they pay, and what choices appear to be available to them.
The system's internal logic was confirmed by Amazon's own executive record. The FTC complaint documents that Jeff Bezos directed executives to accept more defective advertisements (ads that are irrelevant or only marginally relevant to a user's search) as a mechanism for increasing total advertising volume and driving advertising profits. The organic search result, which costs Amazon nothing and delivers the most relevant product to the consumer, was deliberately subordinated to the paid placement, which costs the advertiser and delivers revenue to Amazon. The consumer navigating that environment isn't navigating a neutral marketplace, they're navigating a commercially organized environment whose architecture was designed to maximize platform and advertiser extraction from their purchase moment.
Amazon Dynamic Pricing: The Price Environment
The consumer navigating this commercially organized environment also faces prices that aren't fixed. Amazon's dynamic pricing system makes approximately 2.5 million price changes daily across its catalog, with prices updating as frequently as every ten minutes in fast-moving categories. The system considers demand levels, inventory availability, competitor pricing, time of day, and customer behavior, including product view frequency and cart abandonment patterns. Amazon's own Automate Pricing tool in Seller Central is a disclosed product that enables third-party sellers to participate in the same infrastructure, adjusting their prices algorithmically in response to the same signals.
Amazon has confirmed directly that prices don't vary by individual customer demographics or purchase behavior. A person paying $29.99 for a product is paying the same $29.99 that every other customer in that moment pays. This confirmation is important and accurate, and the paper accepts it as stated.
What it doesn't address is the commercial environment through which that person reached the $29.99 product. That environment (in which products appeared in their search results, sponsored listings appeared above organic results, urgency signals were presented, and alternatives were surfaced or suppressed) was built from their behavioral history and organized by an advertising system paid to intercept them at their highest-intent moment.
The price is the same for everyone. The path to the price is not.
Walmart Connect: The Omnichannel Version
Walmart Connect, Walmart's retail media network, generated $4.4 billion in advertising revenue in fiscal year 2025, up 27% year over year, representing approximately 7% of U.S. digital retail media spending. Walmart doesn't break out exact Walmart Connect revenue as a separate line item but has confirmed the growth rate and the billions-scale revenue from its own earnings calls. Walmart Connect reaches approximately 150 million United States customers every week across its physical stores and digital properties.
Walmart Connect's commercial architecture extends the retail media model into physical space. Walmart operates approximately 170,000 digital screens across its U.S. store network (in self-checkout lanes, on television walls, and in electronics departments) that deliver targeted advertising informed by the same first-party purchase data that powers its online advertising. A consumer who has shopped at Walmart online, participated in the Walmart+ membership program, or interacted with any Walmart digital property has contributed behavioral data to a targeting system that follows them into the physical store.
The Walmart acquisition of Vizio for $2.3 billion, completed December 2024 and documented in Paper Four, connects the connected-home participation surface directly to the purchase-moment interception infrastructure. Walmart now holds the SmartCast platform and its 18 million active account base, a connected television participation surface whose viewing data can be integrated into advertising targeting at the point of purchase. A household that watches television through a Vizio SmartCast device is contributing behavioral data to a system that can intercept a household's members when they walk into a Walmart store or open the Walmart app. The chain from domestic existence to retail transaction is now a disclosed, commercially operational architecture owned by a single company.
Target Roundel: Third Proof
Target's retail media network Roundel generated $649 million in advertising revenue for fiscal year 2024, up 25% year over year, confirmed directly from Target's earnings statement. Target describes Roundel as generating nearly $2 billion in total value when contributions that offset cost of sales and operating expenses are included alongside direct advertising revenue. Roundel reaches 165 million omnichannel guests and operates on first-party data from 100 million Target Circle loyalty members.
Target Roundel's existence and scale establish the argument that is Paper Five's central claim: this isn't an Amazon story. Amazon is the dominant player; its global advertising revenue exceeds all other United States retail media networks combined. But the architecture isn't proprietary to Amazon, it's the operating model of U.S. retail. Target, a mid-market retailer primarily serving American households with annual incomes between $50,000 and $100,000, has built a $649 million advertising business on the same foundation: first-party behavioral data from consumer participation, applied at the moment of purchase to intercept consumers with commercially optimized information, paid for by advertisers who value the precision that behavioral targeting provides.
Beyond Amazon, Walmart, and Target, the practice is expanding across every major retail category: Kroger Precision Marketing operates on the behavioral data of millions of grocery shoppers; CVS Media Exchange operates on pharmacy purchase data and health behavior signals; Instacart Ads operates on grocery purchase intent signals; and Home Depot's retail media network operates on home improvement purchase behavior. By the end of 2024, 64% of U.S. retail executives had plans to add a retail media network to their revenue streams. The retail media network isn't an innovation of a few dominant platforms; the network is an emerging standard business model of American retail. It's a model whose commercial foundation is the behavioral participation data of American consumers, assembled without their recognition or compensation, and monetized at the moment they attempt to make a purchase.
The behavioral profiles assembled across Papers One through Four aren't separate from the retail media architecture documented in this paper. They are its inputs.
A Google search for a product feeds directly into Google Shopping's AI-powered personalized feed, which presents each user with product recommendations and deals calibrated to their individual behavioral profile, picking up where a user left off across sessions. This was disclosed in Google's own product announcement in October 2024 and documented in Paper One. The search behavior that Paper One identified as a participation surface generating $480 per user annually in advertising revenue is the same search behavior that determines which products appear at the top of a Google Shopping feed. The participation and the purchase interception are the same system.
Meta's Generative Ads Recommendation Model (GEM) processes the complete activity history of each user alongside age, location, and behavioral attributes to model what Meta's own engineering team calls the user's "purchase journey," spanning thousands of behavioral events. This was disclosed in Meta's own engineering publication in November 2025 and documented in Paper One. The social media participation that Paper One identified as generating $243 per user annually in advertising revenue is the same participation that determines which commercial offers appear in a Facebook or Instagram feed at the moment a consumer is most likely to purchase. The participation and the purchase interception are the same system.
The Vizio SmartCast platform, whose viewing data Paper Four documented as generating $37.17 in annual Annual Revenue Per User (ARPU) per household, was acquired by Walmart specifically to feed connected television viewing behavior into Walmart Connect's purchase-moment interception infrastructure. The connected home participation surface and the retail purchase interception are now, literally, the same company.
The chain is complete and disclosed. Behavioral participation generates data. Data flows into behavioral profiles. Behavioral profiles are sold to retail media networks. Retail media networks apply those profiles at the moment of purchase to determine what a consumer sees, in what order, with what urgency, from which sellers, at what price point. The consumer who generated the participation at the start of the chain has no recognized standing at any point in it — not at the moment of data generation, not in the advertising transaction, and not in the purchase environment that transaction shapes.
The critics' objection to the variable pricing argument is predictable and accurate in its narrow claim: Amazon has confirmed directly that it doesn't charge different consumers different prices for the same product based on their demographics or purchase behavior. A spokesperson confirmed to Retail Dive: prices fluctuate to meet or beat competitive pricing, not based on a customer's purchase behavior or demographics. This is accepted as stated.
What it doesn't address is the correct claim, which is structural rather than transactional.
The purchase environment is not neutral. It's built from behavioral participation and commercially organized to maximize extraction from the consumer at the moment of purchase. Two consumers searching for the same product on Amazon don't navigate the same commercial environment. One may see primarily organic results reflecting relevance. The other, whose behavioral signals identify them as a high-intent buyer with a demonstrated purchase pattern in a particular category, navigates a commercial environment in which advertisers have paid a premium to appear in front of them at that specific moment because their behavioral data identified them as worth intercepting. The advertisers paid that premium using revenue generated from the behavioral participation documented across Papers One through Four; the same consumer who generated the participation that made them valuable to intercept received nothing.
This isn't price discrimination in the conventional sense; the price is the same. It's something more structurally significant: a commercially organized purchase environment built from the consumer's own behavioral history, optimized to maximize the revenue extracted from their purchase moment by parties whose interests aren't aligned with the consumer's, using data the consumer generated without recognition, compensation, or any standing in the arrangement.
The distinction between this and conventional advertising is the origin of the data. Traditional advertising reached consumers based on the content they were consuming: a television program, a magazine, a search keyword. Retail media reaches consumers based on who they have demonstrated themselves to be across years of behavioral participation: their purchase history, their viewing habits, their domestic routines, their social relationships, their health interests, their financial patterns. The targeting isn't based on context; it's based on the person. And the person who's being targeted is also the person whose participation built the profile that makes targeting possible. They're both the source of the data and the subject of its commercial application, with no recognized role in either.
The four legitimacy conditions established in the Introduction fail in the shopping domain in a way that extends and compounds every failure documented in Papers One through Four.
Survivable Refusal
A consumer who wishes to navigate a retail purchase environment free from behavioral targeting faces a practical impossibility. Amazon controls approximately 38% of all U.S. e-commerce. Walmart reaches 150 million U.S. customers every week. Target serves 165 million omnichannel guests annually. The retail media networks operating across these and dozens of other major retailers constitute the infrastructure of U.S. retail commerce. A consumer who declines to participate in the behavioral targeting infrastructure (by opting out of tracking, declining loyalty programs, and refusing to engage with personalized advertising features) doesn't exit the commercial environment built from their behavioral data, they exit the ability to understand the environment they are navigating. This is because the environment has already been built from their prior participation and will continue to be refined by their current behavior regardless of their stated preferences, because participation in the purchase environment itself generates new signals.
Recognized Standing
The consumer has no recognized standing in any of the commercial transactions that constitute the retail media architecture. When Amazon charges an advertiser to place a sponsored product above an organic result in a consumer's search, the consumer whose behavioral history identified them as worth intercepting is not a party to that transaction. When Walmart Connect sells a brand access to the behavioral data of its 150 million weekly consumers, those consumers aren't parties to the sale. When Target licenses Roundel's audience data to advertisers targeting shoppers across 150 premium publisher sites, the 165 million guests whose behavioral data constitutes that audience aren't parties to the licensing arrangement.
The consumer is the input to every commercial transaction in the retail media architecture. They're a party to none of them.
Transparency of Terms
A consumer entering a retail purchase environment (physical or digital) isn't informed that the commercial environment they are navigating has been organized by an advertising system paid to intercept them using their own behavioral data. The sponsored placements that appear in their search results are labeled as ads, satisfying the minimum disclosure requirement. The behavioral profile that determined which ads appeared, from which advertisers, at what cost to those advertisers, using what data assembled from what participation, is disclosed nowhere in the purchase environment. The consumer navigating Amazon's search results in 2024 was navigating a commercial environment shaped by $56 billion in annual global advertising spending whose inputs included their own participation history. That fact appeared in no disclosure they received at the point of purchase.
Independent Jurisdiction
The FTC v. Amazon antitrust complaint is the most significant available regulatory action against the retail media architecture, and it operates on an antitrust theory (monopoly maintenance, anti-discounting measures, and degradation of search quality) rather than a participation value theory. The complaint doesn't establish any mechanism through which a consumer could claim recognition of their participation as the origin of the commercial environment they navigate. The California Consumer Privacy Act and similar state frameworks provide rights to opt out of behavioral advertising, not rights to compensation for the participation that made behavioral advertising possible. No existing legal framework recognizes the consumer as a productive contributor to the retail media system whose participation generates the commercial value advertisers pay to access.
The four legitimacy conditions fail in the shopping domain as a consequence and extension of every legitimacy failure documented in Papers One through Four. The behavioral profiles that enter the retail media system were assembled under conditions where survivable refusal eroded over time, recognized standing was structurally absent, transparency of terms was legally formalized but functionally absent, and independent jurisdiction addressed privacy harms rather than participation value misclassification. The retail media system doesn't introduce new legitimacy failures. It's the destination to which all prior legitimacy failures led.
In the Origin Economics framework, Y = λ · f(H, K, T) is the formula in which output is a function of human-origin participation, capital, and technology, multiplied by whether the legitimacy conditions of the exchange were satisfied. In the retail media domain, the output is the $55 billion in United States retail media advertising revenue generated in 2024. The human-origin participation is every behavioral signal documented across this series (every search query, every social interaction, every streaming session, every voice command, and every loyalty program transaction) organized by the capital infrastructure of retail media networks and scaled by the artificial intelligence systems that convert behavioral history into purchase-moment targeting. Lambda failed at the origin of every input to this system. It was never present at the destination.
The Personal Data Royalty Formula (PDR) calculation for Paper Five doesn't produce a per-consumer baseline in the way that Papers One through Four produced per-user or per-household figures. The retail media architecture doesn't generate revenue that's directly attributable to a single consumer's participation in the way that platform advertising revenue is attributable to platform users. What it generates is an aggregate measure of what the behavioral profiles assembled from consumer participation are worth to the advertisers who purchase access to them at the point of purchase.
That aggregate measure is $55 billion annually in the United States alone.
Divided across the approximately 260 million American consumers who participate in the consumer economy, $55 billion in retail media advertising represents approximately $212 in annual advertising value extracted from the purchase behavior and behavioral profiles of each American consumer. This figure isn't a per-consumer PDR in the precise sense established in Papers One through Four. It's an aggregate-derived estimate of what each American consumer's participation contributes to the retail media market annually — the value of the commercial environment built from their behavioral history that advertisers pay to access at the moment of purchase.
The Relevance Objection
The primary counterargument holds that behavioral targeting benefits consumers by delivering more relevant advertising: that a consumer who sees ads for products aligned with their interests and purchase history is better served than a consumer who sees random advertising, and that the retail media system therefore provides value to the consumer rather than extracting from them.
This objection correctly identifies that relevant advertising is less wasteful than irrelevant advertising from the consumer's perspective. It fails as a rebuttal to the misclassification argument for the same reason every service-value objection has failed across this series. The question isn't whether the consumer benefits from the relevance of the advertising they receive. The question is whether the consumer has recognized standing as the origin of the behavioral data that makes that advertising relevant, and whether they receive any share of the $55 billion annual United States market that their participation makes possible. The answer to both questions is no. Receiving more relevant advertising isn't compensation for generating the participation that made relevance possible. Instead, it's a description of how the extraction is delivered.
The Competition Objection
A second objection holds that retail media networks create competition among advertisers that benefits consumers through lower prices, that brands competing for visibility in Amazon's search results are competing to offer better deals, and that this competitive dynamic ultimately benefits the consumer.
The FTC's antitrust complaint addresses this objection directly and in the opposite direction. The complaint alleges that Amazon's anti-discounting measures prevent sellers from offering lower prices on other platforms, effectively maintaining price floors across the internet rather than driving prices down. The complaint further alleges that sponsored placements frequently displace lower-priced organic results, meaning the consumer navigating Amazon's search environment often sees a higher-priced sponsored listing before a lower-priced organic result from a seller who did not pay for placement. The competitive dynamic of retail media advertising does not consistently benefit consumers. It consistently benefits the platform and the advertisers with the largest advertising budgets.
The Opt-Out Objection
A third objection holds that consumers can opt out of behavioral advertising through privacy settings, ad preferences, and do-not-track mechanisms, and that the availability of these controls satisfies the consent requirements the PDR framework establishes.
This objection fails on the same three grounds established in Paper Four. The opt-out mechanisms aren't disclosed at the point where the behavioral profile is generated. The default settings across every platform documented in this series generate and retain behavioral data. And opting out of targeted advertising in the present doesn't remove the consumer from the commercial environment already built from their prior participation. The behavioral profile assembled from years of search history, purchase history, viewing behavior, and domestic monitoring continues to inform the commercial environment that consumer navigates regardless of their current opt-out status.
The retail media architecture faces regulatory exposure across two distinct frameworks, neither of which addresses the foundational participation value misclassification this paper documents.
The FTC v. Amazon antitrust complaint is the most significant regulatory action currently proceeding against any retail media network. The complaint's motion to dismiss was denied by Judge John Chun in October 2024, allowing the FTC's Section 2 Sherman Act and Section 5 FTC Act claims to proceed. The complaint's theory of harm focuses on monopoly maintenance, anti-competitive pricing practices, and degradation of the consumer search experience; antitrust theories built on market structure analysis rather than participation value analysis. If the FTC prevails, the remedy would address Amazon's monopoly position and its anti-competitive practices. It would not establish the consumer as a recognized origin of the behavioral participation that makes the retail media system commercially viable, and it wouldn't create any mechanism through which consumers could claim compensation for the participation value Amazon's advertising business depends on.
Privacy regulatory frameworks (the California Consumer Privacy Act, the Colorado Privacy Act, and their equivalents in other states) provide consumers with rights to opt out of the sale of their personal information and to request deletion of their data from retail media targeting systems. These frameworks treat the retail media architecture as a privacy matter: a question of whether consumer data is collected, retained, and used with appropriate consent. They do not treat it as a participation value matter: a question of whether the consumer is recognized as the origin of the behavioral data that constitutes the retail media system's primary commercial asset. The distinction is the same one this series has identified across every regulatory framework it has examined. Privacy frameworks produce the right to limit extraction. Participation value recognition would produce the right to share in the value extraction generates.
The PDR calculation in this paper rests on disclosed revenue figures from Amazon, Walmart Connect, and Target Roundel. It doesn't include the following participation surfaces. These exclusions aren't due to absence of participation, but due to the absence of disclosed, attributable per-consumer revenue at the level required for calculation.
The categories below aren't exhaustive. They're representative of the system.
Grocery and Pharmacy Networks
Kroger Precision Marketing, CVS Media Exchange, Walgreens Advertising Group, and Albertsons Media Collective operate retail media networks on first-party purchase data from grocery and pharmacy transactions. These networks capture the most intimate purchase signals in the consumer economy (food, medication, personal care, and health behavior) without disclosing per-consumer revenue at the level the PDR formula requires. Kroger's loyalty program alone covers approximately 60 million households.
Home Improvement and Specialty Retail
Home Depot's retail media network, Lowe's One Roof Media Network, and Best Buy Ads operate on purchase behavior data from home improvement, appliance, and electronics transactions. These networks reach consumers at high-value purchase moments whose behavioral signals extend into the connected home participation surface documented in Paper Four.
Delivery and Fulfillment Networks
Instacart Ads, DoorDash Ads, and Uber Advertising operate on real-time purchase intent signals (what a consumer is ordering, from where, at what time, and with what frequency) whose commercial precision exceeds most other behavioral signals in the retail media ecosystem. Instacart reaches approximately 8 million active buyers monthly. None disclose per-consumer revenue at the level required for calculation.
Membership and Loyalty Infrastructure
Costco, Sam's Club Member Access Platform, and loyalty program networks operated by airlines, hotel chains, and financial institutions generate purchase behavior profiles whose commercial value is sold to advertisers without per-member revenue disclosure. Sam's Club MAP reported 38% year-over-year growth in 2024. The membership model makes the participation surface more concentrated and more commercially precise than open retail media networks, because the consumer has self-identified and consented to a relationship whose terms don't include recognition of their participation value.
Financial and Payment Layer
Visa, Mastercard, American Express, and PayPal each operate data analytics and advertising businesses built on transaction-level purchase behavior across every retail category documented in this paper and beyond. The payment network sees the complete purchase record (every transaction, every merchant, every amount, every time) across every retail environment the consumer enters. None disclose per-consumer advertising revenue at the level the PDR formula requires. Their full treatment belongs in Paper Ten.
Search and Discovery Within Retail
Pinterest Shopping, TikTok Shop, and YouTube Shopping operate purchase-intent participation surfaces that extend the retail media architecture into social and video environments. Each captures behavioral signals at the intersection of content consumption and commercial intent without disclosing per-user purchase-layer revenue.
In-Store Physical Infrastructure
Digital screen networks in physical retail environments (self-checkout screens, endcap displays, electronic shelf labels, and in-store audio systems) constitute a growing participation surface whose behavioral signals are captured at the moment of physical purchase. U.S. in-store retail media advertising was estimated at $370 million in 2024 and is projected to reach $1.06 billion by 2028. Per-consumer figures are not disclosed.
The $55 billion United States retail media market documented in this paper is derived from the three networks that disclose revenue with sufficient precision for calculation. It's not a measure of what the retail purchase environment extracts from consumer participation annually, but the portion of that extraction that's been disclosed. The floor is not the number.
United States retail media advertising reached approximately $55 billion in 2024. Amazon generated $56.22 billion in global advertising services revenue. Walmart Connect generated $4.4 billion. Target Roundel generated $649 million. Every dollar of that revenue was paid by advertisers to intercept American consumers at the moment of purchase using behavioral profiles assembled from the participation documented across this series. The consumers whose participation made those profiles possible received none of it and aren't recognized as having contributed anything to the system that valued their collective behavioral history at $55 billion annually.
This is where the extraction ends up. Not in a database. Not in a platform's quarterly revenue report. In the commercial environment each consumer navigates every time they search for a product, open an app, walk into a store, or attempt to make a purchase. The environment has been built from their participation. It's been organized to maximize extraction from their purchase moment. It operates at a scale and sophistication that has no precedent in the history of commerce, and the person at the center of it (the person whose behavioral history constitutes its primary input and whose purchase moment constitutes its primary output) has no recognized legal standing anywhere in the arrangement.
The series has now documented five domains of extraction: platform participation, gaming, education, the connected home, and the retail purchase environment. Each paper has added a layer to the same structural argument: participation originates in persons, generates commercially valuable data, flows into systems that don't recognize its origin, and produces revenue for platforms and advertisers who captured the value of that participation while the person who generated it received nothing but the service that acquired them in the first place.
Paper Five closes the loop. The behavioral data assembled from platform use is applied at the moment of platform purchase. The behavioral data assembled from gaming is applied when a player makes an in-game purchase. The behavioral data assembled from educational participation shapes the commercial environment students enter when they graduate. The behavioral data assembled from connected home monitoring follows the household member into every retail environment they enter. And the $55 billion retail media market is what all of it is worth to the advertisers who purchase access to it.
The price was never zero. Paper Five shows what it purchased.
Running total after Paper Five. Add only the lines that apply to you.
Amazon.com, Inc., Q4 and Full Year 2024 Earnings Release, February 6, 2025. ir.aboutamazon.com.
Amazon.com, Inc., Annual Report on Form 10-K, fiscal year ended December 31, 2024, filed with the United States Securities and Exchange Commission, February 2025.
Walmart Inc., FY2025 Earnings Release, February 2025. Walmart Connect revenue figures cited from earnings call disclosures and press releases.
Target Corporation, Q4 and Full Year 2024 Earnings Release, March 5, 2025. Roundel advertising revenue $649 million confirmed from earnings statement.
Target Corporation, Roundel Fact Sheet, 2025. corporate.target.com.
Federal Trade Commission v. Amazon.com, Inc., Case No. 2:23-cv-01495, United States District Court, Western District of Washington. Second Amended Complaint, October 2024. Order Denying Motion to Dismiss, October 7, 2024. ftc.gov.
eMarketer, Retail Media Ad Spending Forecast and Trends H2 2025, November 2025. emarketer.com.
eMarketer, US Retail Media Ad Spending Forecast H1 2024, June 2024. emarketer.com.
Google LLC, The New Google Shopping, Built with AI, Google Blog, October 2024. blog.google. Previously cited in Paper One.
Huayu Li, Xiaoyi Liu, Jade Nie, Ellie Wen, Chunzhi Yang, Jiyan Yang, Nancy Yu, Habiya Beg, Gil Arditi, and Neeraj Bhatia, Meta's Generative Ads Recommendation Model (GEM): The Central Brain Accelerating Ads Recommendation AI Innovation, Engineering at Meta, November 10, 2025. engineering.fb.com. Previously cited in Paper One.
Walmart Inc., Press Release, Walmart Completes Acquisition of Vizio, December 3, 2024. Previously cited in Paper Four.
Imanol Arrieta-Ibarra, Leonard Goff, Diego Jiménez-Hernández, Jaron Lanier, and E. Glen Weyl, Should We Treat Data as Labor? Moving beyond Free, AEA Papers and Proceedings 108, 2018, pp. 38–42.
Eric Posner and E. Glen Weyl, Radical Markets: Uprooting Capitalism and Democracy for a Just Society, Princeton University Press, 2018; reviewed by Peter Isztin in Oeconomia 9(4), 2019.
Shoshana Zuboff, The Age of Surveillance Capitalism, PublicAffairs, 2019; as analyzed in Rosa Aaron Dufva, In Dialogue with Zuboff, Master's thesis, Tampere University, 2022.