As the prominent role of the internet in life generally increased, commerce and financial businesses had to be affected as well. CRM has changed from a system of record to a system of recommendation and random ads are no longer accepted. Nowadays, most of the famous applications and websites use recommendation systems, and the applications or websites that have the better recommender systems are the most popular, they have the ability to attract and retain customers because users like to have a personalized and customized experience. And marketers who use AI in their engines are making a noticeable increase in their revenue. However, there are problems in some of the recommender systems’ algorithms. In this review paper, we present a study about the recommender systems and related works and algorithms proposed to solve the problems and enhance their performance. We made a comparison among previously proposed systems. Moreover, we compare systems’ experimental results which use different technique on customer behavior including statistical methods, and purchase behavior and comparison behavior in calculating users’ similarity. The performance of the algorithms was evaluated by precision and recall values. The comparison shows that the statistical method may show a better performance in some cases.