Research spotlight: N/LAB

In today's digitally driven world, every swipe, click, and purchase we make leaves behind a trail of data. Nowhere is this more prevalent than in supermarkets, where loyalty cards have turned into powerful tools for retailers to understand and cater to our shopping preferences.

The subtle trade-offs we make every time we scan these loyalty cards at the checkout allow us to enjoy discounts and special offers while simultaneously granting supermarkets a front-row seat to our consumer behaviour. But, is there a potential for this data to be used for more than just direct marketing?

A new research project from N/LAB, Nottingham University Business School’s state-of-the-art teaching, data visualisation and research facility, has shown that information from loyalty cards can be used to uncover hidden health insights, such as finding areas with high health deprivation, or highlighting nutrient deficiencies in consumers. It’s all part of the ‘Machine Learning for Good’ ethos that underpins the work at the Lab.

“The sort of digital footprint data we leave behind us - and entrust companies to be guardians of via their loyalty card schemes - is still pretty much untapped”, explained Roberto Mansilla Lobos, one of the team of researchers at N/LAB. “Yes, data can provide valuable insights into consumer behaviour, preferences, and trends, but it can serve purposes beyond pure marketing - we’ve been constantly developing novel AI methods to harness this data to address a wider range of issues, including inequality and sustainability.”

Research at N/LAB focuses on developing AI and statistical methods to gain valuable insights from vast behavioural data sets: by aggregating and modelling large datasets and connecting them to real-world outcomes, it’s possible for researchers to generate findings that would otherwise be unobtainable at scale.

The multi-disciplinary team works closely with businesses and the public sector on releasing social value from digital footprint data, which could then be used to inform data-driven policymaking. Their practice also informs the MSc Business Analytics course, developed for the Business School by N/LAB and taught in collaboration with multinational companies.

“For almost a decade now, the N/LAB has been exploring the potential of transactional data for social good, and we’ve only really scratched the surface”, said Prof James Goulding, Director of N/LAB. “There are so many different applications for this new use of AI and digital footprint data, that it’s hard to cover them all.”

“In addition to simply understanding consumer behaviour better as academics, we’ve had projects on mapping left-behind communities, hidden mental health challenges, loneliness and food insecurity - and in the medical domain, hugely exciting projects that have shown the possibility of early warning of Ovarian Cancer risks or COVID outbreaks through shopping data.”

Their latest work, using anonymised data sets from Co-op loyalty cards, is focused on trying to get a better understanding of public health through consumer behaviour relating to food purchases and their associated nutritional content.

The study showed that analysing purchasing data from loyalty cards could be used to predict deprivation, childhood obesity and future diabetes risks at a local level. For example, the number of calories per pound spent and, to a lesser extent, the proportion spent on cigarettes in an area was found to be an important predictor of high levels of health-related deprivation.

With a high level of predictive power (80%) identified, this kind of data analysis could prove to be a more efficient method of assessing local levels of deprivation than the current Indices of Deprivation (IoD), which are only updated every four to five years.

“There really is huge potential here,” said Roberto. “Our projects with companies such as Co-op have been born out of a shared interest in the potential of how shopping data might better help the communities they serve. Understanding nutritional trends at national scales through grocery purchase behaviour is critical for guiding informed, long-term policies and interventions that can mitigate potential health risks associated with dietary changes and improve the well-being of communities. There is almost too much scope!”

If you're interested in hearing more updates and insights from our leading Business School, you can now subscribe to our Connect Business alumni newsletter.