影響行動廣告有效性之因素研究

Abstract

隨著智慧型手機的普及，導致行動廣告市場快速成長，也即將帶來龐大商機，因此提昇廣告點擊率即成為每個企業所面臨的問題，而點擊率是衡量廣告成效的一個重要指標，也就是當點擊率愈高時獲利率就相對愈高，因此本研究為了幫助企業提昇其獲利能力，欲找出影響廣告點擊率的因素。 透過收集企業實際投放於行動裝置上的廣告資料，以及使用者行為資料，並進一步使用群集分析和決策樹分析，希望能真正找出影響廣告點擊率的因素，而分析結果也發現，廣告本身以及使用者行為皆會影響廣告點擊率，在廣告本身因素中，不同類型的廣告導致不同的點擊情形，而在使用者行為中，像是使用者觀劇類型偏好、使用者觀劇時段偏好也都和廣告點擊有一定程度的相關，除此之外，本研究亦建立了決策樹預測模型，企業除了可以針對不同廣告類型以及使用者行為投放廣告之外，還能搭配預測模型找出高點擊族群，將預算花在刀口上，用最少的預算達到最有效的獲利。

With the ever-increasing number of smart phones, mobile is growing faster than all other digital advertising formats, as advertisers begin allocating dollars to catch the eyes of a growing class of 'mobile-first' users. There is a fertile market for personalized adverting. So, the challenge is how to get your users to click more often on the ads appearing on your mobile property. More users clicking on the ads would primarily mean advertising money for the company. Hence, we collected datas from a company who has its mobile applications and makes effort to enhane their mobile Ad effectiveness. Then we used the clustering techniques and decision tree model to find out what factors are related to the click through rate (CTR). Finally, we found that both Ad itself and user behavior influence the Ad effectiveness. For Ad itself, different types of Ads such as Ad categories cause different CTR. For user behavior, there are many factors that should be considered also, for instance, their drama preference, watching hours preference and so on would cause different CTR too. In conclusion, knowing the key factors and using the prediction model can help company to enhance their Ad effectiveness.