paragraphs 168-175. Even larger publishers (for whom the cost of switching is arguably more significant) have either switched suppliers or renegotiated contract terms to achieve lower CPMs. Contracts also tend to be renegotiated relatively frequently and contract duration is relatively low. This is significant because it suggests that customers consider the threat of switching to be credible. Indeed, if competition is strong and switching is possible, customers will tend to prefer short-term contracts allowing for frequent renegotiations. While a third party claimed that contracts for ad serving tools tended to last about 3 years, the notifying party has provided evidence that the majority of Double Click's contracts do not last more than 2 years ([…]* of DoubleClick's European contracts last up to 2 years – measured by 2006 revenues – and [>70%]* – measured by number of contracts)151. DoubleClick's largest European customers (with more than 1 billion impressions in 2006) have either […]* contracts152. Any significant price rise would therefore likely trigger switching towards competing ad serving tool suppliers or threats to do so as customers have the frequent opportunity to renegotiate prices.
A third-party complainant has also raised the issue that post-merger, the access to CPI by the new entity will create an additional switching cost. As discussed in the discussion regarding the data combination in paragraphs 359- 366, the combination of the Google and DoubleClick customer "databases" is unlikely to provide a considerable additional competitive advantage to the new entity.
The issue of whether price rises in ad serving tools could trigger significant switches to Google's ad network was also considered. Ad serving tools are an input into the delivery of ads and reporting services for publishers and advertisers involved in the direct and the indirect "unbundled" channels. However, as discussed in paragraphs 195-202, the cost of ad serving only represents between 2 and 5% of online advertising costs/revenues for advertisers and publishers153. From the point of view of the publisher, once the decision to use a stand-alone ad serving tool (for direct and/or indirect sales of inventory) has been made, the allocation of the inventory between the various distribution channels will depend on the value at which online space can be sold on each of the channels. The publisher will compare the net profit realized on each channel and distribute its inventory accordingly. The fact that ad serving represents a small fraction of net ad revenues has implications for the likely switching behaviour with respect to ad networks. Indeed, a price increase of 10% of the ad serving tool would lead to minimal variations in the net profit received by the publisher on AdSense's competing networks and this is unlikely to lead to any significant switch towards AdSense. In order to induce any significant switch between ad networks, a substantial relative price increase would be necessary (for instance, a very significant price increase for DFP used on other networks).
See submission of the notifying party's economic advisers of, 26 September 2007, page 10.
See submission of the notifying party's economic advisers of 17 January 2008.
As indicated above, most publishers (and advertisers) responding to the market investigation indicated that the cost of ad serving amounts to 2-5% of their ad revenues/costs (see para. 197 above).