Why the DOJ’s Google Ad Tech Case Matters to You
Lisa Macpherson / Nov 21, 2024Lisa Macpherson is Policy Director at Public Knowledge.
Later this month, the United States Justice Department will offer its closing argument in its antitrust case against Google’s advertising technology business. I have a personal interest in this case because I used to be an advertiser. As a marketing executive in the 1990s and into the 2010s, I leveraged changes in consumer behavior to exploit the precision and cost efficiency of digital advertising. But I left my career in marketing after it became clear that digital advertising capabilities were not only oversold but also causing harm by eroding people’s privacy, compounding challenges for publishers, and facilitating the amplification of harmful content. The DOJ’s case exemplifies how significant Google’s role was in creating and sustaining those harms.
I hope that by sharing how this unfolded, I can help convey why everyone – not just advertisers and publishers – has a personal interest in this case.
A Brief History of Media Buying
Most contemporary advertising – for, say, the last 150 years – amounts to advertisers creating motivating messages about their products and deciding which media channels will distribute them most effectively. Advertisers entered into bilateral negotiations to buy ad space from newspapers and magazines, and eventually radio (in the 1920s) and TV broadcasters (in the 1940s). Advertising agents – who initially worked for publishers, not advertisers – charged newspapers a privately negotiated commission for selling their ad space (and, in many cases, for actually designing and writing the ads). But after accusations from advertisers about rebates and kickbacks to agents inflating their costs, in 1876 ad agency NW Ayer & Son pioneered the "open contract." It was a transparent approach to billing for advertising media that became the industry standard for the next hundred years. Due in part to this transparency, a commission rate of 15% of the advertiser’s investment became the consistent method of compensation throughout the media business. Importantly, this commission was used mostly to pay for the labor of writers and artists who actually created the ads and provided other services, like research, to their clients.
To effectively compete, agencies were expected to understand which media channels would best achieve the advertiser’s objectives. They prepared media plans that reflected the unique ways viewers engaged with print, radio, or TV ads and achieved the advertiser’s goals for how many people saw the ad and how often. Agencies were also expected to know the type of content – the context for the placement of ads – that was most appropriate for the advertiser’s offering. Each media channel had its own methods for verifying that advertisers received the value of their investments. Newspapers’ claimed circulation was verified and recorded in Rowell’s American Newspaper Directory beginning in 1869. Third-party rating services like Nielsen and Arbitron measured and reported on broadcast audience reach. Advertisers were given guarantees of placement and audience delivery, and broadcasters provided “make-goods” that returned to advertisers the difference between estimated and actual viewership of their ads. Agencies provided “clip books” to prove ads had run in the magazines they were scheduled to run in, as well as regular circulation reports comparing planned to achieved reach and frequency.
This was the model of media buying in place when I first started my marketing career. It wasn’t perfect, but it was characterized by willing partnerships between advertisers and publishers or broadcasters; competition among media outlets for the advertiser’s investment; transparent pricing; a predictable take rate for agencies that topped out at 15%; consideration of brand as well as business goals in ad placement; and verifiable delivery of ads across media. Consumers saw ads that were relevant to their wants and needs in appropriate environments and paid prices for goods and services that reflected the plannable, predictable costs of advertising to them.
By the time I reached my first leadership position in marketing, things were starting to change. First and foremost, marketing evolved as a discipline, and chief financial officers began to demand more measurability and accountability for marketing investments. Agencies began to specialize, dividing into “media houses” that bought media in bulk across clients (and earned much lower commissions) and “creative boutiques” that were compensated via a flat fee based on labor costs. With fees from more cost-sensitive advertisers on the decline, media agencies began to consolidate. They had to become more analytical in targeting, measurement, and reporting in order to win and retain clients. Media channels were also fragmenting, creating more competition and choice for advertisers and reducing their costs.
Enter the Internet
Broadscale consumer adoption of the internet in the 1990s offered advertisers an explosion of new insights and new choices. By nature attuned to changes in consumer behavior, we chased our customers in their inexorable migration to the internet for information, entertainment, and socializing. The traditional marketing purchase funnel (referenced by the DOJ in the Google search case), which since the late 18th century had depicted how consumers progress through various stages of consideration before making a purchase, evolved into a “purchase journey,” an interpretation that justified tracking consumers across devices and applications while collecting data about their every move.
Now, theoretically, our media plans could precisely address every stage of this journey; for example, social media to inspire a purchase, display ads to convey product information, and search for feature and price comparisons. (One of Google’s defenses in the ad tech trial was that the DOJ’s market definition was flawed because it ignored “competitors” like Facebook, Microsoft, and Amazon. But those channels do fundamentally different things within an advertiser’s media plan. For example, why advertise on Amazon if you want to invite people to your own website or other retailers? Why advertise on LinkedIn (owned by Microsoft) if you’re not targeting business buyers? These are not substitutable, and I’d have fired a media agency that claimed they were.) Accelerated by pressure from financial officers to leverage their cost efficiency, we moved more and more media dollars from traditional media channels into display ads on the open web – that is, digital ads placed within the content of online publishers. These ads are the focus of the DOJ’s ad tech case against Google.
Media Auctions and the Rise of “Programmatic” Advertising
Early in the digital transformation of marketing, advertisers or their agencies still negotiated directly with online publishers to place ads – like the banner ads frequently placed at the top of websites – in contexts that were relevant to their products. This was generally a real negotiation on an even playing field. The players were known, costs were transparent, commissions were predictable, the context in which the ads would run was known in advance, and make-goods still applied. While banner ads and the like could be annoying, advertisers, publishers, and consumers still enjoyed many of the benefits of the traditional media buying model.
As the number of websites exploded, including hundreds of thousands of new sites reaching niche audiences, we had to automate the process of buying and placing (or “serving”) display ads to publishers. (The DOJ estimates today’s volume of display ads to be 13 billion ads per day). By the mid-1990s, we had learned a whole new language, accompanied by a whole new fee structure. Ad servers delivered and tracked online ads. Ad networks brokered ad buys to leverage unsold inventory between publishers and advertisers. In the early 2000s, ad exchanges came on the scene to feed the twin beasts of speed and efficiency, connecting ad networks and introducing an automated, real-time bidding process to set prices and serve ads in fractions of a second after a user opens a web page. Then advertisers needed access to demand-side platforms, or DSPs, and publishers needed supply-side platforms, or SSPs, to connect to these exchanges. The use of layers of technology (called an “ad tech stack”), algorithms, and the data collected along an individual consumer’s purchase journey to buy and sell ads online came to be known as “programmatic advertising” (that is, those ads that follow you all over the internet).
By the early 2010s, a “negotiation” had transformed into a blind bidding war, conducted in microseconds, ostensibly between advertisers and publishers but facilitated by layers of software programs. Every software provider along the way collected data about the consumer and a fee from the advertiser or publisher – or both. These layers of advertising technology made the ad-buying process incredibly complex and opaque. We couldn’t get answers about how the auctions actually worked, and we could never get our hands on the detailed log files that would show us whether an ad was served and where, who saw it, and how much it cost. We soon learned that our ads could end up on any of thousands upon thousands of small, unknown sites, some of which existed solely for the purpose of hosting ads and many of which were pumping out toxic content. (One report from the U.K. showed that each of the 15 advertisers studied appeared on an average of 40,524 websites.) Now we needed “inclusion lists” and “exclusion lists” of keywords or content where, at least in theory, our advertising could or couldn’t appear. New “brand safety” organizations like NewsGuard (in 2018) and the Global Alliance for Responsible Media (in 2019) sprang up to help advertisers ensure their brands would appear only on quality news sites and avoid supporting propaganda or hoax sites. (Ironically, these services would later come under attack for being part of a “censorship cartel” along with social media companies and the Biden administration – even though these services arose because of insufficient content moderation on the part of platforms.)
During this same time period, open web display ad costs, including fees, were rising sharply. When we asked our advertising agencies – some of the largest and most sophisticated in the world – where this money was going, they couldn’t answer our questions. In 2017, a scathing book called “Bad Men” explained how “the amazing world of ‘ad tech’ magically turns a dollar of online advertising into three cents of value.” In addition to the complex technology fees, bots – software programs designed to repeatedly load webpages – and human click farms artificially inflated the traffic and clicks that were often the basis of payment from advertisers. Some of the most prominent marketing executives in the world began calling this out. In 2017, Marc Pritchard, Chief Brand Officer of Procter & Gamble, the world’s largest advertiser with a $7.5 billion annual advertising budget, described the digital display advertising supply chain as “murky at best, fraudulent at worst.” A year later, Keith Weed, the Chief Marketing Officer of Unilever, at the time the world’s second-largest advertiser, referred to the digital media supply chain as the “digital swamp.” (If these enormous advertisers couldn’t benefit from this system, who could?) In 2020, as the UK report cited above showed, publishers received only 51% of the money spent by advertisers to reach readers, and about 15% of advertisers’ money seems to just… disappear.
So how did we advertisers cover the rising costs of digital media, fees, and software? Often by cutting other consumer communications and services and increasing prices for consumers. It wasn’t until the DOJ’s suit against Google that I came to truly understand what was happening behind the scenes – and why.
DOJ: Google’s Behavior Created the Perfect “Black Box”
By the 2010s, the complexity and profitability of the ad tech stack made it ripe for consolidation. Back in 2008, Google had made the pivotal acquisition of DoubleClick for $3.1 billion, bolting the leading publisher ad server (superior to Google’s own at the time) and an emerging ad exchange to its existing dominance in search text ads. According to the DOJ's suit, Google eventually tied the new publisher ad server to their other services, creating an exclusive pipeline for the advertiser demand coming from their own search and display ad product, Google Ads. Google adjusted pricing through the ad exchange to inflate prices for advertisers and reduce the net to publishers that didn’t deploy Google’s tools. Over time, according to the suit, Google’s policies related to auction bidding, product features, pricing, partnerships, and other aspects of the business evolved to exclusively advantage Google, resulting in higher costs for advertisers and lower net payments to publishers. And, the DOJ contends, Google went on a buying spree, acquiring (and in some cases, then killing) other companies that introduced new innovations or competitive threats in the still-exploding ad tech market. These included products and companies that now make up Google’s supply-side platform, demand-side platform, and analytics and attribution provider, among others. Today, Google’s share of the key services in an advertiser’s ad tech stack, all the way through to publishers, ranges from 40-90%. And remember that 15% media commission that lasted for 100 years? The comparison is imperfect, but the DOJ reported that Google now keeps 30% — and sometimes far more—of each advertising dollar flowing from advertisers to website publishers through Google’s ad tech tools. Publishers have needed to build higher paywalls and charge consumers more for subscriptions, increasing barriers to information.
Part of the benefit of owning the ad tech stack is owning consumer data, and now any insight that advertisers might have received about their customers remained in Google’s walled garden. In fact, according to the DOJ’s report, Google’s determination to master and own the precision-targeted digital ad business and keep advertisers, publishers, and competitors from eroding its dominance led to major compromises in user privacy. For example, before the DoubleClick acquisition, Google’s privacy policies meant it could not combine user data from its own properties (like search, Gmail, and YouTube) with data obtained from other websites. But in 2016, Google changed that policy, integrating user data from across the web into a single user identification for tracking and targeting (some even say, surveilling) consumers across their devices, platforms, and apps. But no one else could access this data. That meant that no one had any real understanding of the consumer’s purchase journey except Google. (We didn’t even know which ads motivated customers to purchase; Google told us that the log files that would have explained how our ads run were proprietary and precluded publishers from sharing them directly.)
As the DOJ summarized, “Advertisers and publishers, the key players in this market, have had scant visibility into the scope and extent of Google’s anticompetitive conduct. As the lone conflicted representative of both buyers and sellers, Google has created a deliberately deceptive black box.” In other words, I wasn’t crazy – there were real reasons for the frustration and confusion advertisers like me experienced in this new Google-owned tech stack.
Google obviously isn’t responsible for all the challenges in the digital ad market, but one would think the company’s dominance of the digital ad supply chain would allow it to round off its sharp edges. Yet, well into the 2020s, it’s been reported that Google still places ads on disinformation sites around the world, often in violation of its own policies. Most recently, the Association of National Advertisers estimated that of the $88 billion in open web programmatic spending, some $22 billion – one in four dollars – is wasteful or unproductive. They also estimated that about 36% or less of those investments actually reach the advertiser’s intended audiences. The ANA noted further that only the largest advertisers have the human, financial, and technology resources to “parse through the complex maze of the digital media ecosystem.”
In my view, if today’s advertisers had more choices, they would never tolerate these problems. And consumers wouldn’t see their brands embedded in – and funding – toxic content or face higher prices.
Google’s Ad Tech Dominance Detrimentally Affects Everyone
I left marketing in 2019, hoping I could use my consumer marketing expertise to teach people how to make better choices about their privacy or use my network to coalesce advertisers to use their own financial power to demand change. I quickly learned that both options faced the same challenge: lack of choice. That is, Google’s monopolistic practices, as described in the DOJ’s suit, mean that neither consumers nor advertisers really have any options. That means they have no leverage to demand lower prices, higher quality, or more data privacy. Public policy interventions to create healthy competition and choice, including through antitrust enforcement and competition policy, plus new solutions to support online news publishers, were a better option. That is what brought me to Public Knowledge.
I hope it’s clear by now why this ad tech suit, like the search case before it, matters to everyone. Google’s dominant control of the advertising technology market has hurt advertisers and publishers in ways that have been demonstrated and quantified. But it also hurts consumers – by imposing on privacy, fueling disinformation, reducing the diversity of viewpoints online, increasing barriers to news, reducing innovation by brands, and increasing prices on goods and services. We need to open up digital markets to bring about more competition and more choice if we want an advertising system that works for the rest of us.
Related reading
Authors
