In this article, an NLP model is suggested to improve e-business performance by identifying and assessing the potential risks. Based on surveying the literature, we found that the customers’ perceptions are not included in developing e-business enterprises. Moreover, a few articles include managing risks based on the indications of users. So, we propose a model which relies on the indications of social media platforms as a fashionable data repository. This model will help managers in building effective e-business enterprises. Moreover, Twitter’s real data for the customers of Amazon Canada is collected and analyzed where the findings refer to the common risks from the point of customers’ view are operational and technological risks. Besides, we provide many performance metrics to confirm the suggested methodology’s efficacy.