Loyalty card data offer a typical example of a contemporary “Big Data” source, allowing compilation of behaviours that inform brand choices, household inventories, promotional impacts and long term behavioural patterns. In addition, customer metadata such as age, gender and postcode are collected, adding a dimension of geo-demographics that can be attributed to transactional behaviours. Despite this, these new forms of data are often produced as a by-product of commercial activities, leading to issues when attempting to apply them in research. Importantly, lack of researcher control over data collection means they may be susceptible to unidentified data error, which is of particular importance when considering spatial applications.
This case study presents the preliminary stages of understanding the potentiality and limitations of loyalty card data for applications within social science research.
Alyson Lloyd and James Cheshire