Adwords Pricing

Years and Authors of Summarized Original Work

• 2007; Bu, Deng, Qi

Problem Definition

The model studied here is the same as that which is first presented in [10] by Varian. For some keyword, $\mathcal{N} = \left \{1,2,\ldots ,N\right \}$ advertisers bid $\mathcal{K} = \left \{1,2,\ldots ,K\right \}$ advertisement slots (K < N) which will be displayed on the search result page from top to bottom. The higher the advertisement is positioned, the more conspicuous it is and the more clicks it receives. Thus for any two slots $k_{1},k_{2} \in \mathcal{K}$, if $k_{1} <k_{2}$, then slot k₁’s click-through rate (CTR) $c_{k_{1}}$ is larger than $c_{k_{2}}$. That is, $c_{1}> c_{2}>\ldots> c_{K}$, from top to bottom, respectively. Moreover, each bidder $i \in \mathcal{N}$ has privately known information, vⁱ, which represents the expected return of per click to bidder i.

According to each bidder i’s submitted bid bⁱ, the auctioneer then decides how to distribute the advertisement...