The European Union has all but announced that it will find that Google has abused its dominant position to undermine competition online – and they are giving the company two weeks to agree to remedies to fix the problems that the EU has identified.
The four main concerns include ones regularly discussed in the media – favoring its own content over potential competitors and scraping content from competitors to bolster its own offerings – but it also includes more specific concerns about how it deals with advertisers.
Since those advertisers are Google’s real customers, that’s significant in getting to the heart of the problem. One concern identified is the way Google structures agreements to lock those advertisers into nearly exclusive deals with Google, thereby shutting out search advertising services.
Shutting down exclusive deals by dominant players is a pretty bread and butter move by antitrust authorities, so that’s not a big innovation for antitrust policy. More significant is what Joaquín Almunia, Vice President of the European Commission responsible for Competition Policy, calls the “restrictions that Google puts to the portability of online search advertising campaigns from its platform AdWords to the platforms of competitors.”
Unpacking that sentence, what this means is that having invested the time to create a Google advertising campaign, Google makes it nearly impossible to create software to easily manage a parallel campaign using the same data on any competing search or online advertising platform.
Why Advertiser Campaign Portability Matters
The problem of Google’s dominance is not just that it controls 85% of global search advertising—and 44.1% of all global Internet advertising.
What’s remarkable is that its competitors, even when they have a moderate share of the market, as with Bing in the United States, make essentially no profits on those market shares. Yahoo, once one of the main competitors to Google in the United States, has largely abandoned search in favor of an alliance where Microsoft’s Bing search engine now powers most search in Yahoo properties—and its payments to Microsoft equaled its costs in 2011.
Even in the U.S. where Microsoft’s Bing search engine has a 27% market share (including its alliance with Yahoo), what is remarkable is that Bing is losing something on the order of $2.6 billion per year in profits for Microsoft. In any other industry, a competitor with the heft of Microsoft — and a track record of pretty vicious competitive actions itself — would itself be seen as a threat but instead is almost not viable economically. In Europe, Bing is barely in the game.
The question is why no competitors can make a profit against Google?
Google’s Monopoly Price on Cost Per Click Advertising
In the case of search advertising, the core product sold is the little text advertisements that appear on search page results after users enter a few keywords. Advertisers bid on those keywords in an auction system, with a combination of the highest bid price and the quality of the advertising site linked to determining which ad gets most prominent placement on the search results page. An innovation in search advertising is that advertisers don’t pay a dime for the ads being listed, but only pay a search engine the price bid in the auction when a user actually clicks on the ad.
This so-called Cost Per Click (CPC) price for a user to click to an advertiser’s site is ultimately the core product being sold. While a dominant player like Google is inevitably going to deliver more clicks based on having more users, all things being equal, the CPC price should be roughly the same since a user ultimately clicking through to an advertisers page should in theory be just as valuable if the customer reaches the page via Google as via AOL or via Bing.
However, even in the United States Google receives an extremely high premium cost per click (CPC). One advertising analyst estimated that the ”average CPC on Bing is somewhere around 1/4 or 1/5 of our average CPC on Google.” Another found that on specific search terms, CPC rates on Bing were slightly higher but still were discounted 49% to 71% compared to Google, while others estimate that the CPC rate is 20% to 40% lower on Bing.
But what this means is that even if an advertiser pays for a search term on both Google and Bing, Bing would end up generating only one-fourth to as low as one-tenth of the revenue Google received from the same advertiser’s campaign using the exact same keywords on each site. Which explains at least part of the reason why Bing may have nearly half the users of Google in the U.S., but generates less than 17% of the revenue that Google receives from search advertising.
This seems like a classic instance of a monopolist being able to charge a monopoly price for the same product as its competitors—in this case a user visiting an advertisers’ site from a particular search engine.
Google’s Barriers to Entry
That this lower revenue has to cover much of the same fixed costs as Google for maintaining a competitive search engine illustrates the first major barrier to entry for existing or potential challengers to Google. Online search engines and related enterprises require incredibly large fixed investments, what law professor Siva Vaidhyanathan described in his The Googlization of Everything as a “monumental collection of physical sites such as research labs, server farms, data networks and sales offices.” Google is estimated to have somewhere on the order of one million computer servers (with other estimates placing that number as high as 1.8 million servers in Jan. 2012) deployed to crawl the web and store the data needed to run its services. While this estimate covers all of Google’s operations, any competitor would have to invest in a significant percentage of those resources to be competitive.
New competitors are therefore automatically at a disadvantage with Google with fewer clicks to begin with, but if they also get a lower price per flick, it’s nearly impossible to meet the fixed costs of running a search engine or any of the other online products Google provides and reach a profit.
What the European Union is de facto arguing is that Google is using anti-competitive methods to prevent successful entry into its advertising markets.
Why Data Portability Could Help Level the Playing Field
In its details, creating an online ad campaign can be incredibly complicated since advertisers not only want to choose a wide variety of keywords and keyword combinations, but also combine them with particular demographic, locational and behavioral attributes of the users to potentially serve up links to different ads for different groups of users (more on that later). Having invested the person-hours in creating such a campaign in Google’s system, an advertiser will find that they cannot simply export the campaign data to recreate the same searches on another platform. Since repeating the costs of designing the campaign twice may not be effective for many companies, especially smaller advertisers, they may simply choose to invest those fixed costs in a campaign at Google where they reach a far larger number of users.
By decreasing the number of advertisers bidding for clicks at competing ad platforms like Bing’s, the CPC price would then end up lower at Google’s competitors. Being the dominant player in an auction-based market, where the costs of competing in multiple markets is high, may thereby deliver a monopoly price for that dominant player.
To give some sense of the costs of managing campaigns across multiple platforms, Google itself through its DoubleClick affiliate provides a tool for advertisers –DoubleClick Search V3 — that will manage campaigns across both Google and Bing, but the price starts at a $10,000 setup fee and costs an additional 2.5% of all advertising spending on a monthly basis.
And here’s the kicker that the European Union is focused on. Google deters third party software from providing similar functionality with tight contract restrictions on third party developers given access to AdCenter’s underlying code (so-called APIs). For example, one provision says that third party software “may not offer a functionality that copies data between Google and a Third Party.” Literally, the contract requires that most data be manually moved between different ad platforms, destroying any efficiency of using a computer to manage multiple platforms.
By requiring manual export of data to alternative ad campaigns, campaigns making repeat, often daily adjustments of ad campaigns will end up with a heavy burden of transferring data. This is “both time-consuming and prone to error” in Harvard Business Professor Ben Edelamn’s words and will “ reinforce the tendency of small to medium-sized advertisers to ‘single-home’ — to use only Google AdWords, to the exclusion of competing platforms.”
And of course, even for companies choosing to run campaigns on multiple advertising plaforms, using Google software means that Google collects a share of advertising revenue on rival sites and will be able to collect data from advertisers on how they use rival advertiser platforms, giving the company advantages in designing their platform to maximize revenue.
So requiring Google to allow third party software makers to manage multiple advertising campaigns on multiple platforms would suddenly make using competitors to Google almost as easy as clicking a button. There would no doubt be some additional costs but far less than currently and most advertising, if the marginal costs were negligible, would prefer to reach all users on the Internet—especially since search advertising is structured such that they only pay if someone actually clicks on their ad and goes to their site.
The European Union action could potentially make a dent in Google’s monopoly position.
What if Data Portability and Other Proposed EU Measures are Not Enough?
However, there is reason to suspect that the Cost Per Click advertising rates for Google AdWords are higher not just because of these structural issues, but because a customer linking to an advertiser’s site from Google is more valuable than a customer linking to it from a competing platform like Microsoft’s AdCenter.
The core source of value being delivered to advertisers by Google is by all accounts how well the company knows its users and that may mean it can deliver customers more likely to purchase an advertiser’s product and, as importantly, helping sell it at the highest price the user may be willing to pay.
The implications for what it means to controls markets through control of user data are quite broad. It also makes better sense of what Google is doing as it spreads its business model into producing smart phone operating systems, email systems, video sites, e-commerce projects and it’s broad range of other endeavors. What unites all of these actions is strengthening the breadth and depth of Google’s database of information on individuals across the Internet in ways that become almost impossible for any competitor to match. To the extent that Google’s higher CPC rates come from its dominance of user data, its monopoly may be becoming almost unassailable.
And as I’ve described in my Cost of Lost Privacy series, its precisely the value advertisers place on knowing invasive amounts about users – particularly who will pay premium prices for their products – that raises the broader and knottier social concerns. With subprime mortgage lenders being the poster children for this kind of “price discrimination” behavioral targeting, untangling the value Google delivers to advertisers and the monopoly position Google has by virtue of its unprecedented database of user information and behavior will require more than the simple laundry list of changes the European Union is currently promoting.
But at least it’s a start and a large statement that there is an antitrust problem with Google’s online dominance.