A Structural Model of Sponsored Search Advertising Auctions
Post date: Dec 19, 2012 5:08:27 PM
Susan Athey and Denis Nekipelov
Sponsored links that appear beside Internet search results on the ma jor search engines are sold using
real-time auctions. Advertisers place standing bids, and each time a user enters a search query, the
search engine holds an auction. Ranks and prices depend on advertiser bids as well as “quality scores”
that vary over time, by advertisement and individual user query. Existing models assume that bids
are customized for a single user query. In practice queries arrive more quickly than advertisers can
change their bids. This paper develops a new model that incorporates these features. In contrast to
prior models, which produce multiplicity of equilibria, we provide sufficient conditions for existence
and uniqueness of equilibria. In addition, we propose a homotopy-based method for computing
equilibria.
We propose a structural econometric model. With sufficient uncertainty in the environment, the
valuations are point-identified; otherwise, we consider bounds on valuations. We develop an estimator
which we show is consistent and asymptotically normal, and we assess the small sample properties
of the estimator using Monte Carlo.
We apply the model to historical data for several search phrases. Our model yields lower implied
valuations and bidder profits than those suggested by the leading alternative model. We find that
bidders have substantial strategic incentives to reduce their expressed demand in order to reduce
the unit prices they pay in the auctions. These incentives are asymmetric, leading to inefficient
allocation, which does not arise in models that ignore uncertainty. Although these are among the
most competitive search phrases, bidders earn substantial profits: values per click are between 40
and 90 percent higher than prices for low-ranked bidders, but are between 90 and 270 percent higher
for high-ranked bidders. Our counterfactual analysis shows that for the search phrases we study,
the auction mechanism used in practice is only slightly less efficient than a Vickrey auction, but
the revenue effects are ambiguous, with Vickrey leading to substantially less revenue for one search
phrase and slightly more for the other. In addition, less efficient quality scoring algorithms can in
some cases raise substantially more revenue with relatively small efficiency consequences.