This paper focus on Influential Ranking Algorithm, because of we find that the current algorithm model has a little shortcoming in recommendation computing process. Then, we give a new improve algorithm model called Influential Ranking Algorithm based on individualized instances. In new algorithm model, we use individualized instances in Common Followers users which show duplication, instead of use the total Common Followers users.We use experiment method to evaluate recommendation capacity between above two model, in Tencent Weibo data. We find that new Influential Ranking Algorithm with individualized instances present higher accuracy in recommendation hit ratio and show less time-consuming in algorithm computing. It can be a new attention in Influential Ranking Algorithm research and Microblog Recommendation Systems. Microblog sites can use the new Influential Ranking Algorithm model to give more influential and appropriate recommendation results to customers, thereby increasing the efficiency of enterprises.