Presentation information

General Session

General Session » [GS] J-3 Data mining

[4B3-J-3] Data mining: contacts with customers

Fri. Jun 7, 2019 2:00 PM - 3:20 PM Room B (2F Main hall B)

Chair:Yoshinori Hjikata Reviewer:Koji Kozaki

2:40 PM - 3:00 PM

[4B3-J-3-03] Maximization of advertisers' pro t in AdNetwork using intrinsic value of conversion

〇Satoshi Kawamoto1 (1. i-mobile Co.,Ltd.)

Keywords:Online Advertising, Optimization, AdNetwork

CPA(Cost Per Acquisition) is a crucial indicator for internet advertising.
Generally, ad serving systems accept target CPA for advertising campaigns, and systems adjust bid prices according to their target CPA.
But, when inputted target CPA are too low, advertisers will not be able to get enough conversions.
In contrast, when target CPA are too high, advertising costs will be too high.
In this study, we aim to construct a model to optimize advetisers' profits.
At first, we expressed the nonliniality between prices and conversions using Poisson regression model.
Next, we derived a model to maximize advertisers' profits.
And then we investigated the behavior of the model by evaluating profits of advertisers and media.