How Google and Facebook Made Digital Advertising Markets Increasingly Opaque to Protect Their Dominance

According to the British Competition and Markets Authority, barriers to entry are so high that not even Amazon is a real threat to Google and Facebook’s dominance. These practices allow digital platforms to raise prices, while advertisers pass the extra costs to their clients. 

 

 


Editor’s note: Following an extensive study, the UK’s Competition and Markets Authority (CMA) expressed concerns in December that Google and Facebook have grown so “large and have such extensive access to data that potential rivals can no longer compete on equal terms.”

 

This is the fourth post in a series on the status of competition in search, social media, and digital advertising markets. It explores the main preliminary findings of a large market study conducted by the CMA, as well as a December 2019 antitrust condemnation of Google in France for abusing its dominant position in the French market for digital ads. The rest of the series can be found here.


 

Previous installments explored the CMA’s preliminary conclusions that the British markets for general search and social media are dominated by Google and Facebook, respectively. For the CMA, this dominance may harm users through reduced innovation, lower quality of services, excessive data collection, or higher prices for goods and services.

 

Both general search and social media are two-sided markets, where an intermediary connects two or more sides (e.g. Google connecting users and websites), such as when Apple connects app developers and iPhone users through its App Store. As such, proper analysis has to consider not only the potentially negative impacts on consumers but also the impact on publishers and advertisers. The Competition and Markets Authority’s study of the dynamics of digital ads, however, indicates that Google and Facebook also control digital advertising markets, where advertisers confront high prices, opacity, and many conflicts of interest.

 

The first question the CMA faced was the extent to which digital advertising competes with legacy media such as TV, radio, or newspapers. The CMA concluded that digital advertising is a market in and of itself. Companies that rely on digital advertising cannot substitute it for traditional media because these legacy channels lack data and the ability to personalize ads and because most online advertisers are smaller companies that do not have the funds or scale required for most legacy ads.

 

The CMA then divided online advertising markets into four non-substitutable segments: search, display advertising (advertisement on banners/videos associated with content), open display (third-party companies selling advertisement inventory on their websites, e.g. newspapers), and classified ads. It analyzed the first three, which correspond to the vast majority of expenditures in the UK.

 

Search Advertisement

 

For the CMA, the body of evidence it analyzed “suggests that Google has substantial market power in search advertising,” the market for ads that search engines display on the top of a search result.

 

Google controls 94 percent of the market for search ads in the UK. Its control over the search advertising market is higher than its control over the search market (where it holds a 90 percent share) because Google charges more per ad than competitors and is more effective at monetizing its platform—as shown by the data gathered by the CMA on the total number of monetized clicks in the UK:

 

 

Google argued that it faces strong competition from vertical search engines (such as Booking.com or Expedia in travel) and from Amazon. After a careful analysis, the CMA concluded that neither threaten Google’s dominance.

 

That is because vertical search engines actually rely on Google for traffic—for example, 50 percent of Booking.com’s operational expenses are with ads it buys from Google. The CMA also concluded that while Amazon might develop into a relevant competitor in the search advertising market, it is unlikely to pose a threat to Google in the near future. Not only is Amazon Google’s main client in terms of search revenue, demonstrating an important vertical link, but the CMA estimates that retail ads (where Amazon could grow) account for only 18 percent of Google’s revenues with ad search, or 33 percent if all the retail sector is considered.

 

Even in a future worst-case scenario, Google would still dominate two-thirds of the market.

 

For the CMA, Google’s dominant position in search advertising is protected by important barriers to entry and expansion of competitors. Initially, competitors face important barriers in developing the user side of search that provides parties with inventory to sell (explored in the second post of this series). Search advertising also has its own important barriers to entry and expansion.

 

First, Google restricts interoperability in these markets. This increases transaction costs in search advertising and prevents smaller advertisers from multi-homing. Second, Google controls much essential proprietary data, in particular tags in third-party websites that allow it to track the success of search ads (whether it led to sales, etc.), key metrics for the market. For example, Google’s tags are present in 88 percent of third-party UK websites, versus 1 percent of Microsoft’s tags. Finally, Google’s vertical integration creates opacity, increases the cost of entry, and enables exploitative practices by Google against advertisers.

 

For the CMA, such practices (which the agency intends to investigate further) may harm consumers through higher prices (as businesses pass-through search costs), diminishing innovation, the charging of hidden fees, and by enabling Google to leverage its dominance in search to other markets. The CMA stressed how Google’s high margins of 44 percent, versus a cost of capital of 9 percent, would be indicative of economic exploitation in this market.

 

Display Advertising

 

The CMA equally concluded that Facebook has market power in display advertising, the market for ads such as videos, banners, or sponsored pages that are displayed in owned and operated platforms (e.g. Facebook, Snapchat, or YouTube ads). 

 

Facebook accounts for almost 50 percent of the display market and is at least four times larger than Google/YouTube, the second major player. If the market is divided in video and non-video ads, something the CMA defends, Facebook’s share grows to almost 60 percent.

 

 

The CMA concluded that Twitter, Snapchat, and other, smaller social media networks do not materially restrict Facebook’s market power as these companies lack scale, sophistication, and supply capacity to challenge Facebook and, to a lesser extent, Google. This highly complex and opaque market is also open to many types of abuses, such as Facebook inflating its statistics on video viewership in the platform (advertisers must rely on Facebook’s own reporting of viewership as the company restricts third-party auditing of the platform). The CMA stresses that it received many complaints reporting fraud in digital advertising metrics and that it will investigate further.

 

While the agency stressed how fraud levels appear to be low and decreasing, it does not believe that reputational concerns alone will solve the opacity and conflicts of interest associated with the fact that the companies that sell ads are also those that report on viewership/effectiveness—largely because Facebook and Google purposefully designed these markets to be opaque and dependent on them.

 

The CMA stressed how Facebook’s privileged position in display advertising is protected by four important barriers to entry. First, the need for scale in the development of platforms, sales teams, bidding technologies, and others. Second, network effects that attract eyeballs in social media markets and generate sales inventory (explored in the third post). Third, Facebook’s control over user data allows it to better personalize ads, increase the reach of campaigns, and prevent other companies from offering competing ads metrics services. Finally, Facebook’s incumbent position in social media allows it to manipulate the market and prevent competitors from reaching enough scale to tackle its dominant position.

 

For the CMA, Facebook’s dominance can negatively impact consumers through a variety of channels. These include higher consumer prices as advertisers pass through increased ad costs in Facebook’s platforms; excessive collection of user data—indeed, the CMA stressed that Facebook has a complex privacy interface and does not allow users to restrict data collection or turn off ad personalization; and increased ad costs as Facebook manipulates the supply of display advertising on its platform.

 

For the CMA, Facebook’s high margins, estimated at 51 percent versus a cost of capital of 9 percent, would be indicative of economic exploitation in this market.

 

Open Display Advertising

 

Finally, the CMA also concluded that Google has a strong position in all segments of the open display market: ads such as videos and banners displayed in smaller scale publisher webpages (e.g. the banners or pop-up ads displayed when you access a blog or a newspaper website).

 

This market is different from display advertising because rather than simply hiring Facebook/Snapchat to do the ad placement, smaller publishers must rely on a complex system of intermediaries that connects them to advertisers (see below).

 

According to the CMA, Google controls 90 percent of the ad server market, leveraging this power to other segments of the market such as demand-side platforms (where it holds a 50-70 percent share) and supply-side platforms (40-60 percent share).

 

 

For the CMA, Google’s dominant position is protected by relevant barriers to entry and expansion. First, Google controls a unique database of personal data that is all but required to match advertisers and publishers in this market. Second, Google’s unique third-party tags (described above) allow it to better track users and report ads metrics without competition. Third, its 90 percent control over the publisher ad service market is protected by high switching costs, as the process of integrating a publisher’s website to an ad server is costly, complex, and may lead to much foregone revenue, forcing publishers to rely on a single supplier. Finally, Google’s vertical integration not only increases the cost of entry but also allows it to leverage its dominance over a key input to other segments, increasing its control over the whole market.

 

Google’s opaque business also created many conflicts of interest and opened market players to all sorts of obscure charges and arbitrariness. Open display is probably the most complex and opaque market in all digital advertising, as it is highly automatized and involves a number of intermediaries between the advertiser and the publisher. For the CMA, advertisers and publishers face a lack of transparency over key aspects of the market, including measurements of effectiveness, quality, how prices are determined, and how much intermediaries are paid—by some estimates, Google and other intermediaries retain as much as 60 percent of advertising expenditures.

 

All in all, the CMA expressed concerns that these problems are also leading to higher prices and lower quality, which are then passed on to consumers. The agency is further investigating many of Google’s leveraging practices. In particular, the CMA stressed how open display markets are essential for the funding of startups and the survival of newspapers, so abuses in market power may have particularly pernicious effects on society.

 

The image below summarizes the CMA’s findings regarding the main forms of consumer harm that may arise as a result of Google’s and Facebook’s dominance of online markets based on advertising:

 

 

Filippo Maria Lancieri is a fellow at the Stigler Center for the Study of the Economy and the State. He is also a JSD Candidate at The University of Chicago Law School.

 

The ProMarket blog is dedicated to discussing how competition tends to be subverted by special interests. The posts represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty. For more information, please visit ProMarket Blog Policy.