Simulating ad performance on PPC platforms

Abstract

Online advertising is one of the most widely used forms of advertising in recent years. One form of online advertising is search engine advertising, i.e., paid links displayed on the website of search engine operators. Google, the world’s leading search engine, processes over 3.5 million search queries per minute, and the trend is rising. So a well-placed ad can reach many potential new customers. However, due to the black-box nature of Google, it is not easy to analyse information about an advertising campaign. This work proposes a system that simulates the advertising campaign on the Google Ads platform, Google’s advertising service. In this way, an advertiser gets insights into the characteristics and statistics of campaigns without placing them on Google, which saves costs and time. Also, with the system, it is possible to learn the optimal parameters based on the campaign. Then, the advertiser can use the parameters to improve the performance of a Google campaign. We perform various tests to verify the functionality of our system and present use cases that an advertiser can perform with the system.