Predicting customer purchase monetary with advertising / Hui-Hsin Huang

This paper constructs a stochastic model to describe customer’s online shopping spending and e-commerce effects. This paper focuses on the reflection of advertising effect which can directly predict customer spending. It assumes that customer purchase monetary can be composed by adverting volumes and its lagged effect which is the duration that customer is still interested to this product after exposing the advertising. Based on these two variables, it can be calculated the probability distribution and the expected value of purchase monetary. The empirical data which is from an online shopping site of women clothes, bags and shoes is demonstrated to estimate the parameters of the proposed model. When the customers browse the web page, and the popup ad of relative product will show up. The unit of observation time is a month (30 days). It can obtain the data of advertising effect from the volumes of popup ad when this customer starts to browse the site and the data of spending amount. It shows the proposed model has good fitness with empirical data. This result can be applied to company profit management through measuring the ad effects to predict customer monetary spending