In the traditional research of customer survey, the basic research methods are interview or questionnaire. For the relevant data obtained, researchers can use a variety of analysis methods for comparative analysis, so as to get important research conclusions. At this stage, with the improvement of online platform and technology level, there are many online reviews of products. The main reason for the popularity of online reviews is that they can contain a lot of customer expectation information and customer opinion requirements. If we summarize, sort out, compare and analyze these comments, we can get important analysis results, which can fully show the needs of customers and play a very important reference role in decision-making. However, a lot of research on this aspect, for the prediction of customer demand is still not deep enough, there is a lot of room for improvement. Based on this development status, this paper develops and designs an online review content mining system based on opinion mining and fuzzy time series method. Opinion mining can effectively mine the important customer needs hidden behind the online review content, as well as the different weights of various customer needs. Fuzzy time series method can predict the future importance weights based on the mined information. At the same time, a case study is taken to compare the effectiveness and feasibility of this method with exponential smoothing method, simple moving average method and fuzzy moving average method in the online review analysis of electric iron. It is found that the method studied and designed in this paper has higher accuracy and reliability.