An innovative bidding optimization model for online advertising
This article is part of the Academic Alibaba series and is taken from the paper entitled “Optimized Cost per Click in Taobao Display Advertising” by Han Zhu, Junqi Jin, Chang Tan, Fei Pan, Yifan Zeng, Han Li and Kun Gai, accepted by the 2017 Conference of the Association for Computing Machinery’s Special Interest Group on Knowledge Discovery and Data Mining. The full paper can be read here.
Acknowledged by The Economist as China’s largest online marketplace, Alibaba’s Taobao operates a state-of-the-art online advertising system. Alimama, Alibaba’s advertising platform, recently revealed details of how a new cost per click method has been designed to transform outdated traditional ad pricing benchmarks.
Dubbed Optimized Cost-per-click (OCPC), this method treats ROI-based arrangement of ads as an internal optimization issue, while the overall user experience and platform revenue is treated as external ones. Overall, OCPC provides an innovative new approach that minimizes user and business sacrifices, all while providing better returns to advertising partners.
Traditional fixed-bid advertising methods such as cost per mille (CPM) and cost per click (CPC) are coarse-grained volume distribution and matching modes, and have been unable to evolve with online display advertising.
Due to the static nature of traditional digital ad formats, they are only able to target specific, pre-determined user groups through rigid sets of prices. This results in ineffective resource allocation, both on the platform end when it comes to assigning ad space, and for advertisers, who are now starting to demand fine-grained price bidding, volume distribution and matching.
Though Google AdWords’ eCPC (effective cost per click) method adjusts advertisers’ bids based on users’ potential conversion rate, it currently does not directly optimize for many key indicators of the Taobao ecosystem.
The current advertising system suffers from two flaws. Firstly, rigid ad pricing is economically inefficient when adapted to continually fluctuating volumes. Secondly, traditional eCPM (effective cost per mille) profit maximization arrangements myopically focus on short-term targets, rather than also accounting for changes in user experience, business volume and other variables, thus making them unsuitable for ensuring the long-term health and prosperity of an E-commerce platform.
The full paper can be read here.