Segmentation of the PAYE Anytime Users

Abstract

PAYE anytime is a web application designed and implemented by the Office of the Revenue Commissioners of Ireland. The application allows the Pay As You Earn (PAYE) customers in Ireland to manage most of their tax affairs online. By using easily accessible technology, PAYE customers can update their information and process most of their tax credits and reliefs online in a clear and effective manner. This online system was designed and implemented in order to reduce the volume of direct contacts between Revenue and its PAYE customers, to decrease costs and improve overall efficiency and effectiveness within the organisation. Moreover, the usage of such an e‑channel allows Revenue to record important information that can be analysed with the aim of improving overall customer service. Therefore, management of this strategic contact channel is paramount to Revenue’s continued advancement and improvement of its online services. This paper describes a segmentation of PAYE anytime users. The segmentation was conducted to understand the profiles and behaviours of these customers. This unsupervised data mining method produces an unbiased, self directed portrait of PAYE anytime customers. The data analysed were extracted from the weblogs of the PAYE anytime online application, which contains information about the users’ navigation. The data were linked to the users’ information held in the Revenue data warehouse in order to access all recorded details about PAYE anytime users. This information consists of the tax credits claimed, the value of tax credits, time period and similar attributes. By linking the online behaviour with the users’ information and mapping on the demographic details of the users, it was possible to identify the different segments and their profiles. The results of this segmentation improve Revenues understanding of the PAYE customer base. Knowledge gained with this project can be applied in a number of areas. Naturally, the profiles and behaviours associated with each segment can be strategically used for customer intelligence policies, allowing specific services to be tailored around customer profiles. Moreover, the analysis can point to improvements of the design and structure of future iterations of the PAYE anytime application.