USA And Canada Ads Spend Model Establishing And Analysis Based On Adwords Paid Advertising Product Of A Certain Website

AdWords, an advertising product of a well-known website, is a quick and easy advertising partner for the customers to promote their products on the Internet. Every day there is a large number of users searching on the Internet by using the website when the customers’ ads will be showed in the search results by AdWords. AdWords hopes through marketing activities to attract new customers. From the marketing point of view, one enterprise promoting its products is no longer just for the fame, but hoping to get more immediate benefits. They struggle for an urgent marketing demand better than effective marketing channel. And with the development of Internet, the amount of data, data processing technology and the ability to deal with the data have got a big process, which largely changes people’s various aspects like the way of thinking, business marketing model, etc. Therefore, the application of Customer Relationship Management (CRM) and data mining tools are essential to maximize the profits. The purpose of this study is to develop a potential customers scoring model of AdWords, that is, according to the advertising spend as the basis of marketing targets, in order to better understand the characteristics of potential customers. Target marketing can reduce cost and improve the efficiency of the market. According to the definition of potential customers of the United States and Canada by AdWords, with large number of the company’s advertising spend values missing, models are respectively set up by using the data of two countries and to explore the characters of potential customers.Therefore, this article first reviews the current condition of Customer Relationship Management (CRM) and data mining, introduces several kinds of commonly used data mining tools and statistics technology, discusses the modeling steps and data processing method, and then combine Logistic regression and linear regression with scoring model to deal with customer’s data.This paper includes research work and conclusions as follow:1) In order to improve response rate and the quality of responders, establish a two phase response-spending model to get the expectations of the companies’ advertising spend;2) Logistic model is established based on the obtained data in order to get the characteristics of the potential customers, and find out that representative potential customers characteristics variables is highly related to a company’s industry, the model effect is significant,that can make a marketing activities to achieve twice the result with half of the effort;3) Finally, provide the basis and opinions for AdWords marketing team, and pick out the potential customers according to the model result, as the basis for the marketing targets.