Faced with various online ads, firms are hard to choose the most appropriate advertising channels which have best advertising effects. Online advertising has immediate and carry-over effects. We constructed a comprehensive evaluation model of multi-channel online advertising effects which can evaluate not only immediate effect but also carry-over effect based on lag effect factors. Then, we conducted a restricted grid search and multiple linear regressions to estimate the immediate effect and carry-over effect of paid search ads, mobile phone message ads and e-mail ads based on user behavior data and transaction data of an e-commerce website. The results show that the immediate effect intensity of paid-search ads is the highest, the carry-over effect duration of e-mail ads is the longest, and the cumulative carry-over effect intensity of e-mail ads is the highest. This study puts forward suggestions on how to evaluate the effects of multi-channel online ads more accurately, which can guide this e-commerce website to make better advertising strategy for online marketing.