Forecasting Customer Lifetime Value

As the number of companies fighting for customer attention grows exponentially, the cost of customer acquisition rises and the retention processes become more complex. Even though it becomes easier to obtain prominent amounts of data, it is simultaneously harder to distinguish the data that may serve a specific business objective. The progress in web analytics has caused rapid growth of the volume and variety of data focused around the customer and its relationship with businesses. Firms are looking for strategic value that can be generated through big data analysis, the kind of value that helps to adapt to the ever-changing business environment and gain competitive advantages. The main aim of this paper is to fill the existing gap in e-commerce behavioural analytics by utilising a number of predictive data analysis methods and introducing some measures (based on obtained information) that may aid in efficiently growing business online. The research and analysis are carried out on real data of a regional e-commerce retailer of FMCG sector. The results of the research are intended to be of benefit to the company and assist in finding out what may accelerate customer retention and improve the relationships with clients in general. Likewise, the results may be of value to the marketers’ community, providing a concept about which metrics and means of predictive analysis should be used to determine the needs of online customers and ways of online business improvement. The first chapter of the paper focuses on the current state of the FMCG sector in e-commerce and expands on the concept of Customer Lifetime Value and why it is important for a healthy growing business. The second chapter reviews customer behaviour in relation to e-commerce, describing more extensively some insights related to customer behaviour and elaborates on the process of calculating and predicting CLV. The most valuable metrics of online customer behaviour are described, as well as main takeaways from predicting future CLV indicators. The obstacles and limitations that appear when analysing customer behaviour are highlighted. The third chapter merges the obtained information of the two previous chapters together and offers suggestions and marketing solutions to improve business performance in the future. The last chapter provides an insight into the research that was conducted for an existing e-commerce website and describes how predictive modelling was used to determine behavioural patterns among e-commerce customers. In the final chapter, the recommendations for the business are developed based on research results, as well as plans for future research are included.