International Think-Tank on Innovation and Competition
Yahoo! Economics
March 25th, 2008
We are at a crucial point of the development of the market for online advertising, and, after the recent merger between Google and DoubleClick, the possible acquisition of Yahoo! by Microsoft could entirely reshape the structure of this market. It is calculated that worldwide spending on advertising is currently above 600 billions $, of which at least 40 billions are spent in the online field, less than 10 %. Since 1994, when HotWire sold the first banner for advertising, and 1995, when Infoseek introduced search-based advertising, online ads have been constantly growing in all of their different forms (search advertising associated with search engines, display advertising, classified listings on web sites and email advertising).
The market is clearly destined to increase its share in the advertising market, at least for the following reasons: 1) the Internet is rapidly growing and the large majority of websites generate revenues from advertising (with the notable exception of transaction websites as eBay); 2) other devices as mobiles and TV will be always more often connected to the Internet; 3) software innovation allows more efficient mechanisms to reach targeted consumers, today on the basis of the characteristics of search (keyword bidding system) and of the same websites (contextual advertising), in the near future on the basis of the characteristics of the Internet users as well.
A merger between Microsoft and Yahoo! would produce substantial synergies between two companies active in complementary markets and, most of all, could create a strong competitor against Google, whose dominance in pay-per-click Internet advertising has been now combined with DoubleClick's dominance in ad serving services. It is now interesting to evaluate the consequences for the future of online advertising of these two crucial mergers (one completed, one possible in the near future). First of all, when does a merger involving a market leader hurt consumers? A pre-condition for this is that the merger relaxes competition between two firms and leads them to increase prices. Two further requirements are needed: first, the merger doesn’t create efficiency gains; second, it doesn’t attract endogenous entry of new firms. Of course, the extent to which a merger leads to price increases depends on multiple factors. First, high substitutability between the products of the merging firms brings higher incentives to increase post-merger prices (after all, firms producing unrelated goods would have no reason to increase their prices). Second, when price-cost margins are high, it is more profitable to increase prices. Third, when the firms operate in a multi-sided market (as a software platform which charges both publishers and advertisers), a price change can optimize network effects between sides leading to higher profitability.
A merger between Microsoft and Yahoo! would mainly create synergies in R&D efforts without consequences on the prices of the main products of Microsoft (its operating system and Office) and Yahoo! (Internet services), which are complements and not substitutes. Antitrust authorities would hardly contest such a merger. Moreover, it would allow Microsoft and Yahoo! to join their forces and develop search engine capabilities and online advertising services able to represent an alternative to Google, so as to strengthen competition in the Internet. For these reasons, also the stock market appears to appreciate the offer of Microsoft.
Let us look at Google now. As well known, Google is the leading search engine in the world, with 53 % search traffic in US at the beginning of 2008, against 17% of Yahoo!, 7 % of Microsoft’s Live, 6 % of AOL and 3% of Ask, but with even higher market shares in other parts of the world (with the exception of Japan, where Yahoo! is the leader, and China, where Baidu is the leader). By the way, Yahoo! has been the leader in the US until 2002: until 2000 it was followed by Altavista, in 2001 by Microsoft and in 2002 by Google, which subsequently obtained the lead. Notice that competition for the market is crucial since access to any search engine is free and simple, and most users simply employ the search engine that is regarded as the most valuable. Of course, network effects are crucial here and, in the absence of substantial product differentiation, they lead to a single dominant player: today, Google enjoys such a position. Beyond this, Google dominates the lucrative business of placing text ads next to search engine result. Google AdWords (launched in 2000) accounts for about 70 % of search advertising revenue worldwide. DoubleClick leads the industry in directly placing banner ads on third-party publishers, accounting for more than 75 % of the direct (or reserved) channel, that is the valuable ad inventory that large web publishers directly negotiate with the advertisers (through their direct sale forces). Of course, a lot of the advertising space available on large websites and all of the space available on medium size and small websites cannot be sold in direct negotiations. Therefore, most of online advertising is typically sold through indirect intermediaries that buy the so-called “remnant” ad inventory from publishers and sell it to advertisers. This can be seen as a separate market from the market for the direct channel.
Google and DoubleClick play a major role in the market for intermediation services for remnant ad inventory. Google provides vertically integrated intermediation platform between online web publishers and advertisers: Google's AdSense reaches more than 80 % of the ad revenue in the indirect channel with integrated ad networks. The Google platform targets advertising to the relevant websites (so-called “contextual advertising”) and pays the web publishers with a percentage of its revenues. Meanwhile advertisers buy inventories from the platform through Vickrey auctions on the keywords that match the content of the webpages and lead Internet users to click on their advertisement: charges are typically for each click on the ad (CPC, cost per click), and the highest bid for each keyword association wins (with the price given by the second highest bid). DoubleClick offers an ad serving and ad management product, DART, for publishers (DFP) and advertisers (DFA). Such a publisher tool manages the inventory of a website, receives the ads from ad networks and delivers them in the relevant inventory (according to the behavioral history of Internet users), usually at a fixed cost per thousand impressions (so-called CPM) which is a small percentage of the price that the web publisher charges on the advertisers. The market share of ad revenue served by DoubleClick’s DFP in the indirect channel with non-integrated ad networks is around 75 %. Since almost 60% of online advertising taking place through the indirect channel adopts integrated intermediation, Google controls about half of the market and DoubleClick about a third of it. After the merger they are going to control at least 80 % of the worldwide market for online advertising.
The two intermediation services that were separately offered by Google and DoubleClick before the merger were highly substitutable and, as a matter of fact, many web publishers used both for different inventories in their websites and in different moments. Needless to say, for these publishers the two services were interchangeable: they could easily recode some space on the website served by one channel to be served by the other channel. Adoption of the publisher tool provided by DoubleClick, or switching to a different one involve high sunk costs in terms of substantial investments in software, in training the staff, coding all of the publisher’s web pages, creating novel datasets, transferring ad campaigns to the system and so on, with all the associated business risk. For the same reason, multi-homing (with multiple non-integrated ad networks) is highly inefficient in this case. Notice that the high switching costs, together with the difficulty of building alternative high quality intermediation services in the short run represent a substantial barrier to entry of new firms in the short and medium run, which is the relevant time frame in such a rapidly evolving market. Finally, notice that this merger could create efficiencies in R&D spending, but it is not going to create efficiencies through reduction of the marginal costs of production for the simple reason that these marginal costs (as typical of the software market) are already close to zero. Consequently, any increase in mark ups after the merger is going to lead to an increase in prices.
The bottom line of this story is quite simple. Before the merger, competitive forces kept online advertising prices under control: DoubleClick could not increase prices because many consumers would have quickly switched toward AdSense, and Google could not increase prices because many customers would have switched immediately to DoubleClick’s products. After the merger, these competitive constraints are about to disappear: Google will easily increase the price of DFP services being sure that most of the lost customers would simply switch to AdSense (the direct channel would be unavailable and more expensive, and other publisher tools would be penalized by the high switching costs). Given the high margins and the network effects that Google could enjoy by increasing its market share, the profitability of the price increase would be further enhanced. Moreover, Google could exploit the merger in its favor and against the competitors in other ways. For instance, DoubleClick uses a sophisticated algorithm to match ads to Web pages, and reports to advertisers, to inform them whether their ads were seen by the customers they tried to reach; now Google could shift the algorithm, favoring Google's own AdSense rather than other competitors. Finally, higher prices of DoubleClick may jeopardize the competitiveness of other products competing with the Google’s integrated channel, and the dominance of Google after the merger would be strengthened once again.
At this point, the outcome of the attempt of Microsoft to buy Yahoo! is crucial for the destiny of the market for online advertising: a merger with Yahoo! could create the conditions to compete effectively with the dominant force of Google in online advertising. In such a case final consumers could only gain, as always when competition gets more aggressive.
.
