Research On New Energy Vehicle Product Satisfaction Improvement Strategies Based On Online Reviews

In 2022,China’s new energy annual sales exceeded 6.8 million units and the market share increased to 25.6%,gradually entering a period of full market expansion.With the intensification of competition in the new energy vehicle market and the changing needs of consumers in the market,companies should respond quickly to consumer feedback and improve consu mer satisfaction with their products.Traditional satisfaction survey methods(questionnaires and telephone interviews,etc.)cannot obtain timely and accurate feedback from consumers.With the continuous development of the Internet,more and more people choose to express their opinions and views on products in the Internet,and online reviews are considered as a reliable source of business intelligence.How to use online reviews to help new energy vehicle manufacturers improve consumers’ satisfaction with their products is the focus of this paper.Research on product improvement based on online reviews has mainly focused on how to identify the improvement priority of product attributes from online reviews,but has not further clarified the level of improvement,nor has it considered how to design improvement strategies to maximize the improved satisfaction under cost constraints.To address the deficiencies of existing research,this paper proposes a method to determine the improvement strategy of new energy vehicle product satisfaction considering the nonlinear relationship between the performance of product attributes and satisfaction,taking the online reviews of Auto Home as the research object.The research is divided into the following two parts:First,the online reviews of Auto Home were analyzed,and the data were obtained and pre-processed by crawler software.Natural Language Processing(NLP)was used for sentiment analysis,and Possibility Distribution-based Hesitant Fuzzy Linguistic Term Sets(PD-HFLTS)was used to code the data to characterize individual satisfaction of the product attributes.A new filtering rule was created to eliminate invalid online reviews,and individual satisfaction was aggregated based on the reliability of online reviews to obtain the satisfaction of product attributes.The satisfaction function of consumer satisfaction with the product attributes was determined by estimating the parameter values using the non-linear curve fitting method with the satisfaction model proposed by the existing studies using the AH-100 vehicle rating system as the performance of the product attributes.Second,a questionnaire survey was used to collect information on consumers’ preferences for product attributes under different usage scenarios,and Best-WorstMethod(BWM)was used to calculate respondents’ product attribute preferences,and product attribute preferences were obtained through information aggregation.Then a cost factor is introduced to describe the cost function of product attribute performance.The nonlinear programming model with the objective of maximizing overall vehicle satisfaction is established to solve the target performance of product attributes,and the improvement strategy of product attributes is formulated.In this paper,we analyze the data of the Xiaopeng P7 pure electric vehicle as a case study,and the results show that when no additional resources are invested,in order to maximize the overall satisfaction,the manufacturer of Xiaopeng should appropriately reduce the performance of the attributes "driving experience"," handling" and "comfort" which have a low degree of overall satisfaction.while focusing on improving the performance of "space" and " endurance" attributes.And furthermore,the degree of improvement of the product attributes is clearly given.By conducting sensitivity analysis on the resource variable in the model,the performance changes of product attributes with different resource costs are clarified.