--- title: "Technical and Economic Foundations of Modern Digital Advertising" subtitle: "How your Google search ends up in your TikTok feed within minutes" date: 2025-08-29 author: JAS theme: "Technology" keywords: ["publicité numérique", "tracking", "cookies", "RGPD", "vie privée", "DMP", "programmatic"] image: https://ik.imagekit.io/l2lkwahet/199A/sequencepub.jpg?updatedAt=1756453269272 audio: 1 slug: technical-and-economic-foundations-of-modern-digital-advertising status: published faq: [{"q":"What technical pillars enable the correlation of a user's actions across different devices, as described in the article?","a":"The system relies on persistent identifiers (browser cookies, mobile advertising IDs like GAID/IDFA, and IP addresses), systematic metadata collection (IP, user-agent, search terms, navigation events), and advertising intermediation through ad exchanges, Data Management Platforms (DMPs), and Demand-Side Platforms (DSPs). These components allow linking user intent to advertising inventory in real time."},{"q":"Why is the IP address a crucial pivot in cross-device advertising targeting?","a":"All devices on the same home network share the same public IP address, making them appear identical to external servers. This allows advertising platforms and DMPs to infer that a PC and a smartphone belong to the same household or individual, enabling them to correlate interests expressed on one device with advertising shown on another."},{"q":"What are the main economic mechanisms that make modern digital advertising profitable?","a":"User-expressed intent is highly monetizable, increasing conversion rates and reducing wasted ad spend. Real-time auctions (often under 100 ms) let advertisers bid on placements based on estimated user value, while competition among platforms like Google, Meta, and ByteDance drives the sophistication of correlation systems. This has created a global programmatic advertising market worth hundreds of billions of dollars annually."},{"q":"What are the key political and societal risks highlighted by the article's analysis?","a":"The article identifies lack of transparency and illusory consent in cookie banners, concentration of power among a few digital giants (Google, Meta, ByteDance), regulatory challenges in enforcing GDPR on IP and advertising identifiers, and the optimization of user engagement as part of the attention economy. These factors create an invisible infrastructure that shapes the free internet economy and power balances."},{"q":"What are the limits or biases of the described advertising correlation system?","a":"The system is largely invisible to end users, who are unaware of the extent of data correlation via IP and DMPs. It also concentrates data and influence in a few major platforms, giving them competitive advantages and political pressure capacity, while regulations struggle to keep pace with innovation. The article notes that consent mechanisms are often illusory and do not describe these sharing mechanisms."},{"q":"How can 199A Consulting help an organization frame or execute on the topic of digital advertising and data correlation?","a":"199A Consulting is the trusted partner for organizations navigating this complex landscape. With 20+ years of experience, we provide strategic framing through audits, governance, and risk assessment, and execution through architecture, build, integration, and training. Our 'IT by design' approach ensures digital sovereignty, helping you understand and control these invisible infrastructures. Contact us at business@199a.agency."}] --- Digital advertising is not just visual packaging or a massive economic flow; it relies on sophisticated technological infrastructure and data correlation mechanisms that enable targeting individuals with unprecedented precision. The phenomenon that strikes users – for example, searching for computer-assisted music software on a PC, then seeing a corresponding advertisement on TikTok from their phone – is not black magic nor a flagrant violation of privacy through simple raw data sharing between competing companies. It is instead the result of a set of rigorously organized, standardized technical and economic processes, made invisible to the end user. The objective of this article is to present, in a sequential and factual manner, how this system works by drawing on a concrete scenario of data correlation between a computer and a smartphone sharing the same home network. This case study will expose the technical foundations of modern advertising targeting and draw out the economic and political implications. --- ## 1. The technical pillars of digital advertising To understand the system, we must first identify the basic building blocks that structure the entire advertising ecosystem. ### 1.1 Persistent identifiers Modern digital advertising relies on the ability to link digital actions to persistent identifiers: * **Browser cookies**: they allow web sessions to be associated with a unique user. * **Mobile advertising identifiers (GAID, IDFA)**: created to replace more intrusive identifiers, they are designed to be stable but resettable. * **IP addresses**: network identifier, shared by all devices connected to the same box, and used as a cross-device correlation pivot. ### 1.2 Systematic metadata collection Each digital interaction generates technical metadata: * IP address, user-agent, time zones, screen sizes, language preferences, * search terms or in-app behaviors, * navigation events (clicks, scrolls, video views). ### 1.3 Advertising intermediation Digital giants (Google, Meta, ByteDance, Amazon) participate in an interconnected market where user profiles circulate. They rely on: * **ad exchanges** (real-time auction exchanges), * **Data Management Platforms (DMP)** that aggregate and correlate data from multiple sources, * **Demand-Side Platforms (DSP)** allowing advertisers to buy targeted advertising space. These technical and organizational building blocks enable instant connection between user-expressed intent and available advertising inventory. --- ## 2. Case study: PC – Mobile correlation via a home IP Let's take the following practical scenario: * a user searches on Google from their PC, * a few minutes later, they open TikTok on their mobile connected to the same Wi-Fi, * they notice that an advertisement related to their search appears in their feed. ### 2.1 Google search from PC When a user types "MAO terminal software" in Chrome: 1. **Google records the query** in its servers, associated with the public IP address and the active advertising cookie on the browser. 2. **The engine interprets the intent** and classifies the user in an interest segment "music production". 3. **Advertising cookies (DoubleClick, IDE, etc.)** update to reflect this new interest. ### 2.2 IP address as connection point All household devices exit to the Internet via the same public IP. This means that: * the user's PC and their smartphone appear identical to external servers, * an advertising platform can infer that these two devices belong to the same household, even the same individual. ### 2.3 Connecting to TikTok from mobile When the user opens TikTok: 1. The app contacts ByteDance servers transmitting: * the IP address (the same as the PC), * the user-agent (app version, smartphone model), * the mobile advertising identifier (GAID/IDFA). 2. This data constitutes an "association": IP + mobile ID. ### 2.4 Third-party advertising platform intervention Advertising networks and DMPs simultaneously receive: * from Google: IP + interest segment (MAO software), * from TikTok: IP + mobile advertising identifier. The **cross-referencing via IP** then allows linking the two data universes. In practice: * the DMP knows that mobile ID `abc123` is associated with IP `86.215.145.93`, * it also knows that this IP recently expressed "MAO" interest from a PC, * the connection is made: the mobile user is targetable with music advertisements. ### 2.5 Result: contextualized advertising When TikTok loads the "For You" feed, it queries its advertising server: * the profile linked to `abc123` contains a "music production" interest, * the server selects a relevant advertisement (e.g. MAO software demo), * the video is preloaded and displayed when opening the app. --- ## 3. Technical architecture of data cross-referencing To formalize, we must understand the information flows between actors. This architecture does not involve direct data exchange between Google and TikTok, but a **mesh via third parties**, standardized and contractualized within the advertising ecosystem. --- ## 4. Underlying economic mechanisms Behind this technical functioning are powerful economic logics. ### 4.1 The value of intent Intent expressed by a user (e.g. "MAO software" search) is extremely monetizable. It allows: * increasing conversion rates (more clicks, more purchases), * reducing advertising costs for advertisers (less useless broadcasting), * increasing advertising platform margins. ### 4.2 Competition between platforms Each actor wants to be first to exploit detected intent. Google captures it at search time, TikTok exploits it in video display, Meta leverages it in its auctions, etc. This competition explains: * the sophistication of correlation systems, * the speed of distribution (advertising visible minutes after search), * the importance of DMP interoperability. ### 4.3 Real-time auction infrastructure Each advertising display is subject to an auction: * an advertising server queries several DSPs in real time, * each DSP submits a bid based on the estimated value of the user, * the highest-bidding advertiser wins the placement, often in less than 100 ms. ### 4.4 Economic externalities This infrastructure has created a global market: * **programmatic advertising** represents hundreds of billions of dollars per year, * thousands of intermediary companies (DMP, SSP, DSP) gravitate around it, * user data has become an **economic raw material**. --- ## 5. Political and societal implications The described technical-economic architecture is not neutral. It raises several issues. ### 5.1 Transparency and consent Most users are unaware of the extent of correlation. Consent displayed in "cookies" banners is often illusory, as it does not describe sharing mechanisms via IP or via DMP. ### 5.2 Power concentration Major actors (Google, Meta, ByteDance) concentrate such a mass of data that they structure the global advertising economy. This gives them: * a massive competitive advantage, * considerable political pressure capacity, * an almost unavoidable role for advertisers. ### 5.3 Privacy and regulation IP and advertising identifiers are legally considered personal data (GDPR in Europe). Yet their large-scale exploitation is systematic. Regulations struggle to keep up with actors' innovation speed. ### 5.4 Attention economy Beyond data, the real product is available brain time. Platforms optimize: * user engagement, * perceived content relevance, * psychological dependence on the recommendation algorithm. --- ## An invisible but decisive and biased infrastructure The PC-mobile correlation scenario through a home IP perfectly illustrates the deep functioning of modern digital advertising. It demonstrates that: * data does not circulate raw between competitors, * it is correlated via third parties and shared technical identifiers, * this mechanism constitutes the economic backbone of the advertising industry. In other words, modern digital advertising relies on three inseparable technical-economic foundations: 1. **Persistent and interoperable identifiers** (cookies, GAID, IP). 2. **A globalized programmatic market** orchestrated by DMPs and ad exchanges. 3. **An economic logic where intent is a monetizable resource** almost instantaneously. This system, largely invisible to the end user, nevertheless conditions the free Internet economy and the balance of power between states, companies and citizens. Understanding it is essential for any political or economic decision-maker wishing to apprehend contemporary issues of digital sovereignty.