Google has achieved a level of digital integration where, as host Andrew Couts outlines in the WIRED segment "Incognito Mode," it is nearly impossible to go online without interacting with Google in some way. While most users accept that Google funds its services—from Gmail and YouTube to Search and Maps—through advertising, Couts meticulously reveals the breathtaking scale of this data reliance. In 2024, Google generated a monumental $264 billion from ad sales, accounting for nearly 75% of its total revenue.
This powerful financial engine is fueled by an ecosystem that harvests an extensive range of personal data, including location, search history, online activity, music preferences, ad interactions, and even potentially health information. This collected data, Couts explains, empowers Google’s ad technology to ensure advertisers know "exactly what you want to buy" and, critically, "exactly when you're most likely to buy it". The company’s market dominance provides this leverage: Chrome holds 67.9% of the worldwide browser market share, Android commands over 71% of the global OS market share, and Google Search maintains nearly 90% of the online search market.
Related article - Uphorial Radio
To illustrate the sheer volume of data involved, Couts advises examining the developer privacy disclosure, which lists categories such as location, contact info, browsing and search history, identifiers, analytics (like purchases), and user contacts. He contrasts this heavily with competitors like the privacy-centric Brave browser, which reports collecting only usage data and identifiers such as a username. Couts divides Google’s collection methods into two primary types: active inputs, which are conscious submissions, such as Chrome browsing data or map searches, and passive collection, which often occurs in the background via location tracking, cookies, and fingerprinting. Cookies, which Couts describes as "little bits of data that websites download under your browser," are foundational to this passive tracking. While some cookies are necessary for basic functionality, Couts focuses on advertising cookies, which track "a bunch of information about what you do online," including preferences, specific web page interactions, and movement, all across the internet. Due to Google's status as the world's largest advertising company, its persistent cookies are encountered virtually everywhere online, regardless of the browser used. Couts notes that cookies have historically been the "primary way" that companies have "fueled the surveillance economy".
Significantly, Google reversed its policy at the end of 2024, choosing to continue allowing third-party cookies in Chrome while simultaneously experimenting with new, allegedly less invasive tracking methods. One such experimental method, according to Couts's analysis, is fingerprinting. Fingerprinting collects data about the user’s device—including the operating system, device model, and even the rate of battery drain—to combine this information into a unique ID. Couts views this technique as a "more nefarious form of tracking" than cookies because, unlike cookies, which can be deleted, altering the physical device used to access the internet every time is "impossible". Couts draws attention to Google’s major policy change in February 2025, which shifted from banning fingerprinting to apparently allowing the practice, suggesting this may be the company’s new strategy for online tracking beyond cookies. Addressing a common public misconception, Couts clarifies that Google claims it does not sell user data outright, unlike data brokers. Instead, it sells access to targeted audiences via its ad platforms through a sophisticated real-time bidding process. This system allows advertisers using complex algorithms to rapidly bid on ad placements, targeting highly specific "segments or buckets" defined by complex characteristics, such as income, profession (e.g., in tech), age (e.g., middle-aged), and vehicle type. Despite Google’s insistence that this data is anonymized and that advertisers do not receive direct access to individual user profiles, Couts highlights studies showing that it is "almost trivial to use anonymized data to identify individuals".
The risk inherent in such a vast concentration of information is, as Couts observes, amplified in the era of AI, especially with the introduction of generative platforms like Gemini. While Google asserts it uses only publicly available information (like social media posts or web articles) to train its large language models, Couts warns that there is currently "no law in the US that prevents it from changing how it's going to use that data in the future," leaving the door open for future use of sensitive information. Couts recounts the significant legal fallout from Google’s data practices, including settlements and lawsuits for deceiving consumers about location tracking (involving DC and three other states), a class action lawsuit for collecting sensitive health information, and various settlements concerning facial data collected via Google Photos.
Acknowledging that total avoidance is "almost impossible"—given that a 2019 investigation found popular apps like Spotify and Uber "break when you block access to Google servers"—Couts provides actionable privacy steps. These include switching from Chrome to privacy-centric browsers like Brave, attempting to avoid Google-branded apps where possible, and, most importantly, changing settings within Google services to disable personalized ads, turn off location history, and delete browsing data. Couts concludes with the pragmatic realization that since Google’s collection is constant, stopping the use of its products will make the user "irrelevant to the company very quickly," cementing the enduring rule of thumb: "if a product is free, that probably means you're the product".
