Research On Evaluating The Feature Fatigue's Effect On Customer Equip

Nowadays, the competition of business is becoming keener and keener. To maintain and attract enough customers, enterprises begin to integrate large amount of functions to their products. As the products growing multifunctional, the enterprises notice that there is another serious problem in front of them, the product feature fatigue. Feature fatigue is generated by three factors: too many leaving-unused functions, high-level of using difficulty and bad function performance. All there three factors would lead to customer complain and return purchase, and then the customers of the enterprise would be lost. The enterprise's customer equip would be inevitable badly hurt if feature fatigue keeps going serious, and feature fatigue has become a problem that enterprises have to face. So, the traditional product design way, which concerns adding functions too much, has to be guided and optimized urgently. This paper, based on the former researches, deeply studies the effect from feature fatigue to customer equip (abbr. FF-CE effect). The research work can be summarized as follows:First, to dealing with the problem of evaluate the FF-CE effect for the products on sale, a method with direction is proposed. According to the literature and research background, we decide the structure variables, and raise the relationship hypothesis. Design questionnaire to get the observation variables which belong to different structure variables. Then build the structure equation model under the hypothesis and variables. After that, the hypothesis would be tested, and the result would be used to fix the quantity of FF-CE effect. An instance of Samsung S5830 is given to practice this structure equation model based FF-CE effect evaluating approach. Second, to dealing with the problem of anticipate the FF-CE effect for the products in design stage, an artificial neural network based approach with direction is proposed. In this part, we novelty propose a new concept, meta-product. Meta-product is a system of index that would completely define a product. The approach is begun with the laying expression of product function, and then turn the product information is turned into Boolean information. According to the Boolean information, we could get the product's meta-function information. Determine the structure of the neural network by repeatedly try work. Use training set as input data to train the network. The trained network could be used to anticipate the FF-CE effect of products in design stage. An instance of fictitious product is provided to practice this approach.