Precision Marketing Method Based On Consumer Journey

With the penetration of Internet technology in various fields,online shopping has become a mainstream shopping method.The massive amount of user data brought by online shopping has transformed advertising into a direction of precise placement.The diversification of consumer demand has made precision marketing an important means of achieving advantage within fierce market competition.Precision marketing refers to the use of various available means to provide the right information to the right user,at the right place,at the right time,through the right channel.This implies precision in four areas including user,content,channel and timing.At present,the research in precision marketing field has been mainly focused on only three of the four areas,which are user,content and channel.Data mining and machine learning methods are also widely used in the research.However,the research on precision of timing is rarely seen.This paper applies the concept of consumer journey and consumer decision process model as the framework for approaching the timing problem of marketing.This paper utilized grounded theory,decision tree,and survival analysis in developing a comprehensive marketing method that can solve the timing problem.First,the study collected and analyzed the latest clothing shopping experience of 40 consumers through in-depth interviews.Through work sessions with experts,the research identified four types journey with significantly different behavioral characteristics.These were labeled roaming,flexible task,rigid task and commuting journeys.The interview content was then pro-grammatically coded through a grounded theory approach in order to summarize the factors that influence consumer decisions,resulting in two main categories of personal characteristics and shopping context,respectively.Second,based on the results of in-depth interviews and grounded theory,a questionnaire about consumers’ most recent apparel shopping process was developed to obtain data that could quantitatively describe complete consumer journeys.A predictive modeling process was then developed using decision tree algorithms and survival analysis,which predicts the length of the window when consumers are receptive to marketing information.The modeling process wasdivided into three steps.(1)Product information demand level(probability)calculation.Calculation of the probability of product information demand at each stage of the journey mode through in-depth interviews and questionnaire research.(2)Consumer journey type classification model.The decision tree C4.5 algorithm is used to create a classification model with input variables of demographic characteristics,shopping context and shopping needs,and output which type of journey mode the consumer is most likely to adopt.(3)Consumer shopping stage survival probability model.With the decision tree C4.5 algorithm,the shopping behavior is used to predict the consumer’s current shopping stage,given that the consumer’s journey mode category is known.Then,with the Cox proportional risk model,the probability that the consumer remains active at the current stage over time is constructed using shopping behavior as a co-variate.Third,this research proposed a marketing approach that can solve the timing problem of marketing based on the above prediction process,such as a communication plan that include the start and end time of the message delivery based on predicted window of consumer receptivity to marketing messages.This approach can avoid ineffective reach and enhance message relevance along the temporary dimension,which eventually lead to improved the efficiency of precision marketing.