would also have to purchase DFA. Given the small proportion of their total expenditure on AdWords accounted for by AFS, advertisers would be unlikely to switch from DFA's rival products to enable their ads to continue appearing on AFS. The bundling strategy limited to AFS would thus likely result in a large degree of "opt-out" from AFS and a subsequent loss of revenues from ads that would have been placed by AFS180.
In the light of these figures and circumstances, the merged entity would be unlikely to risk losing even only a few customers in its core business of search advertising and (search) ad intermediation, where the vast majority of its revenues are earned and where the revenues from each large customer are high, in an attempt to force its low-margin ad serving products upon those larger customers.
Secondly, in the online advertising environment transactions often involve customised solutions or services that are uniquely priced. Bundling is usually an attractive and profitable strategy in order to discern customer's willingness to pay in a context where prices are posted and uniform across customers. In the online advertising industry, bundling would not enable the new entity to increase profits because prices are highly individualized. On the advertiser side, both Google through its auction mechanism for keywords and DoubleClick through its direct negotiation with customers have the ability to vary the price of their products according to customers' willingness to pay. Similarly, on the publisher side, both Google through its negotiations with its direct partners (which account for around [>80%]* of its AFS revenue) and DoubleClick through its direct negotiation with publishers, have the ability to vary the price of their products according to publishers' preferences. In such context, one of the attractions of bundling usually disappears.
To sum up, the large losses that the described bundling strategy would likely produce in the merged entity's core business, coupled with the limited gains from revenues in ad serving would render any such strategy unprofitable to the effect that the merged entity would lack the economic incentive to engage in such practice.
Including non-search intermediation in the bundle
As regards the possible extension of the bundle so as to include also non-
search intermediation, the incentives may be different as revenues achieved by Google through non-search intermediation are much more significant than the revenues that DoubleClick achieves through the sale of its ad serving technology. As a consequence, the merged entity could more easily compensate losses of search customers which refuse to accept the bundle by gains from non- search intermediation customers181. However, as regards the bundling of
A similar argument can be made for AFC (Google's intermediation services for the sale of contextual ads)
as also AFC accounts for only [<10%]* of advertiser spending on AdWords and most advertisers currently also choose not to opt out of AFC because AFC is available at no significant extra cost.
The market investigation indicated that the commission that intermediators receive for their (non-search) intermediation services on average amounts up to approximately 25% of the total price of the relevant display or contextual) advertisement.