Publius Project
Essays & conversations about constitutional moments on the Net collected by the Berkman Center.

Money for Provision and Clicks for Free

Essay by Francois Lévêque, January 22, 2009

When I was kindly invited to write a short essay on Advertising and Internet for the Publius Project, Google’s encyclopedia and potential Wikipedia killer, Knol, was not launched, yet. So, I did not have to balance between the reputation to be one of the first European scholar to contribute to the Berkman Center’s new forum and the money from ads incrusted on my post on Knol.Google.com. A hard choice for any economist like me! Anyway, the recent initiative of the firm from Mountain View provides me with a good case to illustrate a few patterns on ads and Internet.

Online advertising will have a profound impact on the future of the Internet

Online advertising has boomed and become the fastest growing segment of advertising. In terms of dollars spent, it grew at 40% annually between 2003 and 2007. In the future, advertisers are expected to spend $54 billion on online advertising in 2011, that is, the double of last year. One reason for this huge growth is that advertising can be better targeted online than through radio, TV or newspapers. As is well known, the Internet is a very effective invisible mechanism to collect information on consumers and track their behavior. Targeting and profiling raise serious privacy concerns but also provide some benefits. Firstly, it provides more relevant information to users. Do not forget that 40% of searches are commercially motivated. Secondly, it reduces inefficiencies raised when you put ads under the eyes of people who are not interested in them (e.g., a TV spot on shampoo for bald people). My point here is not to assess whether for society those benefits offset the harms on privacy. (Honestly, I do not know) I just want to point out that there are strong forces that, like it or not, make money water-shedding in online advertising.

As those dollars move online, all Internet giants see a chance to strengthen their position in online advertising business. Google, outbidding Microsoft, bought DoubleClick, a software company specialized in serving ad displays online for $ 3.1 billion. Microsoft took over aQuantitative, involved, inter alia, in ad networking business and (like DoubleClick) in ad-serving technology for $6 billion. Yahoo! bought RightMedia, the leading ad exchange, and BlueLithium, a behavioral targeting network, for $680 million. AOL made also a series of acquisitions for an undisclosed amount. In a nutshell, the Internet giants want all to become intermediaries between the advertisers and the websites.

No doubt, such a large change will impact the future of the economy and the industrial organization of the Internet. The example of Knol and Wikipedia provides a concrete illustration:

Unlike Wikipedia, Knol offers contributors the opportunity to make money. If they accept ad displays or text ads intertwined with the piece of knowledge they wrote, they get a share of the revenues Google will receive from the advertisers. The more popular their entry in the Google encyclopedia, the more revenues they’ll get. Some users of Wikipedia – those who value ads - and certain contributors of Wikipedia– those who are attracted by money - might shift to Google’s encyclopedia. Today, nobody knows the extent of these defections, if any, and therefore nobody is able to forecast what will happen. The predictions are even diverging between insiders. For the former Wikimedia foundation chair, Florence Devouard, Knol is the strongest threat Wikipedia has ever been confronted with whereas the honorary Chairman of the foundation, Jimmy Wales, has welcomed the entry of Google in declaring that “The more good free content, the better for the world”.

In the worst case scenario, Knol will eliminate Wikipedia. Less users for the latter means less donors, less donors means less money to operate and to renew the network of servers, insufficient network capacity leads to lowering the quality for users, and the degradation of quality results in reducing the number of users. Less users may lower the attractiveness of Wikipedia to contributors, and in turn the decrease in number and in quality of contributors adversely impacts on the number of users.

By contrast, we can imagine that Google will not catch up and Wikipedia will be stimulated. [Today, Knol provides less than 150.000 entries, especially in the field of health and medicine, against 2.5 millions for Wikipedia.] Competition in online encyclopedia industry like elsewhere could boost people and organizations. Google’s entry into this space can result in better quality of Wikipedia content, speedier access for its readers and an easier process for its contributors.. The important point here is that whatever the scenario, there will be an impact on Wikipedia: the introduction of online advertising changes the future of the Internet.

Free provision of content will decline

The Internet is a unique experience of both free access and free provision. Wikipedia is an iconic illustration of these two free-of-charge sides. There is a myriad of others, such as websites delivering online academic contents or blogs without ads. For economists, the magic does not lie with free access but with free provision. We are familiar with what we call two-sided markets, where you can deliver one product for free (e.g., radio programs to households, payment card to individuals) because you sell another product to other clients (e.g., audience to advertisers, payment facilities to shopkeepers). The economic puzzle is why people decide to incur costs (at least the costs of their time and their efforts) without a financial counterpart. Hence, the tons of economic papers on the motivations of Open Source Software developers. Fun, altruism, ideological convictions, reputation, learning, personal marketing are part of the story.

Motivations are often complementary. Consequently, more money flowing into online advertising means a decreasing share of content that is provided completely for free. For instance, some contributors to collaborative networks or certain bloggers will consider the money from ads as the cherry on the cake. They will use it as ‘argent de poche’. Others will invest more time to improve the quality of their writings. Some other individuals might decide to enter into content provision expecting the ads will remunerate their time.

The issue is to what extent this second trend will affect the content on the Internet. A major concern is online advertising may reduce diversity and free speech. It is especially shared by those who view non-financial and financial motivations as opposed and as leading to different outcomes. In our example, contributors who want to maximize their ad revenues would rather write an entry dealing with what to do in Singapore rather than with the reign of Ramses VIII (1129-1126 BC). It is likely there will be more entries on Knol on tourist cities than on Pharaon dynasties. Interestingly, some registered Knol authors chose not to accept ads on their posts. Unfortunately, Google did not disclose the percentage that those advertising-adverse contributors represent. If it is significant, it would be possible in the future to demonstrate whether topics addressed in contributions with ads are different from those without ads. A good experiment to test how advertising influences content.

In France, there is currently an interesting debate on whether the complete elimination of advertising on public TV channels will result in more cultural (and less trash) programs. The example of BBC One, the UK channel, shows that this expectation might be misleading. The impact of advertising on free speech is also controversial. On the one hand, non-politically correct statements may be refrained and self-censured to avoid the risk of a decrease in ads revenue. On the other hand, as a complementary incentive, money through ads may increase expression of opinions. According to Google’s chief legal officer, David Drummond, “We […] found that online advertising promotes freer, more robust, and more diverse speech. […] Without [it, bloggers] would not be able to dedicate as much time and attention to their publications as they do today.”

No doubt free provision will decline— but we will only get a clearer picture in a few years of the subsequent impact on online content.

Antitrust control and litigation will expand

Web-based industries are young and innovative, two features that are generally associated with strong competition. However, because of scale economies and network effects, leading Internet firms enjoy high market shares. Google, for instance, with more than 650 million unique visitors per month worldwide (including 150 millions in the US), enjoys about 70% share of the search traffic (53% for the US only). Such levels of marketshare inevitably attract the scrutiny of antitrust enforcement agencies, either through investigations or through complaints by rivals and consumers. Moreover, the official green light is required for a merger. The current consolidation in the web-businesses gives antitrust agencies more data and facts regarding how the competition process is going in several markets. At first glance, online advertising markets seems immune, as the number of both online content providers and advertisers is huge and therefore concentration is low. However, both rely on a small number of firms to make the process work (e.g., selecting and serving the ads on the websites, measuring the impact). The creation of a dominant gateway connecting advertisers and online contents producers has been one of the main concerns raised by Microsoft to try to prohibit the acquisition of DoubleClick by Google. Neither the US, nor the EU antitrust authorities retained the argument. Both gave a full clearance to the merger. Of course, a high market share is not anti competitive per se. It is generally the outcome of successful business strategies that lead to better products and services to consumers. Even when high market shares enable the extraction of monopoly profits, companies need not be hindered by antitrust law, as such high profits can incentivize firms to innovate. Moreover, as is well known, competition in several web-businesses takes the form of a race where the winner takes all and then is challenged and displaced with time. The antitrust concerns mainly arise when the winner undertakes anti competitive practices to maintain or strengthen its dominant position. Imagine, for instance, if Google modifies the algorithms of its search engine to facilitate the high ranking of Knol entries on the page results-- less users will be directed to Wikipedia and more to Google’s encyclopedia. Today, about 40% of the consultations of Wikipedia come via queries treated by Google search engine. This figure demonstrates how Wikipedia would be if Knol pages were systematically ranked above its own. It is likely that, should Google attempt to o leverage its leadership on the search market to get market share and ads revenues on online encyclopedia publishing, it would infringe antitrust law. Note that the more Google activities spill over into the content business, the more the suspicion that its search engine is biased will arise. Google currently serves ads to users primarily via search results pages; it is obviously attractive to try to continue to serve ads to them when they arrive at the desired website. However, Google is only allowed to succeed its expansion in publishing and media by merit, that is, because its contents would be better for users, not because it drove them to click on its pages in trafficking PageRank. As far as the Publius Project is concerned, do not worry: Google would never deliver a more terrific content even if contributors get some dollars from ads! François Lévêque is professor at the Ecole des Mines, Paris and a Visiting Professor at the Faculty of Law at UC Berkeley. His research, teaching and consulting interests are in the areas of antitrust, intellectual property rights and network regulation. He blogs here.