中文摘要：

消费者对新产品的持续性动态采用过程可以看作有限状态自动机的状态转移（待购状态、在购状态、休眠状态）．在综合大众传播、口碑效应、首次购买、重复购买、顾客流失、以及季节性扰动等因素的情况下，运用有限状态自动机构建出符合动态采用模式的市场成长模型．此模型更具综合性和精确性，不但可以涵盖经典的Bass模型和后续发展的服务增长模型与尝试一重购模型，还可以推断出以往难以估计的重购系数和流失系数．实证检验快消品、耐用品和服务等品类数据，结论表明新模型比现有三种模型具有更好拟合性，对未来销售的预测力也更高．该模型有利于深入了解新产品在社会系统中一般性的扩散和成长规律．

英文摘要：

The dynamic consumption process could be considered as state transition of finite state automata. Trial purchasing, word-of-mouth effect, repeat purchasing, customer attrition, and seas,）nality can influence the market growth for new product simultaneously. Combing all these factors, we develop a generalized market growth model based on finite state automata. This model is more relaxed and rich than the Bass model, service growth model, or trial-repeat model. Findings from five categories in China show that the forecasting effective- ness is greater than those of other three models. Our research could help firms to understand deeply the diffu- sion and growth patterns for new products and guide decision makings on marketing strategies.