Nottingham University Researchers Develop AI-Powered Tools to Visualise Food Impact Data
Researchers at the University of Nottingham have developed AI-powered tools to investigate the environmental impact of food consumption. Through the interactive web portal, users can visualise the environmental footprint of their own food shopping choices. Notably, they have developed a new, objective indicator of environmental food purchasing behaviour using store card data and machine learning algorithms. Joe Duncan-Duggal, Foodsteps’ Chief Scientific Officer, sits on the project’s advisory panel. Drawing on Foodsteps' extensive expertise in understanding and communicating environmental impact data, Joe provides input to project direction to ensure it is relevant and impactful for the food industry and food policy.
Consumers hold a range of beliefs about the environmental impact of food. When common impact misperceptions are corrected, many consumers indicate that they will behave differently. Until now, there has been insufficient data to assess these claims, however establishing their veracity is key to designing demand-side implementations to reduce the environmental impacts of the food we consume.
Public misperceptions of environmental food choices
Environmental sustainability may not consistently rank as the primary driver of food choice, yet it remains a meaningful consideration for many consumers. When other things are equal, most people opt for the sustainable option. However, evidence suggests that consumers do not accurately understand which food choices are environmentally preferable.
Substantial misperceptions regarding the environmental impact of everyday food products persist, and these inaccuracies are likely to influence decision-making. Correcting such misunderstandings, therefore, represents a potentially powerful lever for shifting consumption patterns. Recent research conducted by Fletcher and colleagues (2025) at the University of Nottingham offers important new insights. Notably, this study expands the scope of previous work by examining perceptions across a wide range of supermarket products rather than focusing solely on raw ingredients or niche categories such as organic foods. By integrating recent advances in environmental data mapping, the researchers were also able to systematically compare consumer perceptions with robust scientific estimates of environmental impact.
Heuristics in environmental judgments
Findings indicate that most people use heuristics, or cognitive shortcuts, when evaluating the environmental impact of food. Two dominant heuristics emerge: whether a product is plant- or animal-based, and the extent to which it is processed. While the assumption about plant and animal products is a relatively reliable guide, the perceived importance of processing is often misplaced. Many highly processed foods are assumed to carry a substantial environmental burden, yet this is not necessarily the case. Common items such as crisps, bread, and soup – despite their classification as highly processed – can have comparatively low environmental footprints . This disconnect underscores a broader tendency to conflate a food’s nutrition with its environmental impact.
At the same time, certain environmental dimensions remain largely overlooked. Factors such as water stress and water pollution rarely feature in consumer evaluations. As a result, the environmental impact of products like nuts and rice is frequently underestimated, despite their significant water-related footprints.
Misunderstanding the scale of impacts
While general patterns of understanding exist for certain food categories, the scale of impacts within broad categories appear difficult to comprehend. Beef, for instance, is widely recognised as environmentally intensive. However, consumers tend not to differentiate sufficiently between beef and other meats. In reality, the disparity is considerable: producing beef mince can generate approximately ten times the greenhouse gas emissions associated with chicken(1). Such gaps in understanding suggest that even when consumers are directionally correct, they may underestimate the magnitude of differences between options.
The behavioural potential of correcting misperceptions
A particularly compelling aspect of the Nottingham study lies in its exploration of how correcting misperceptions could influence behaviour. After participants were presented with scientifically grounded environmental impact data, many reported surprise – especially in cases where their prior assumptions had been highly inaccurate. Importantly, participants also indicated an intention to adjust their purchasing behaviour accordingly: increasing consumption of lower-impact products and reducing consumption of higher-impact ones. Despite this sentiment, participant responses may merely reflect social desirability rather than genuine intent. Only by observing subsequent purchasing behaviour can we confirm the validity of these responses.
Food footprint calculator and donate your data
The research team aims to fill this gap by analysing purchase card data as an objective measure of behaviour. The team has developed an interactive web portal that allows individuals to visualise the environmental impact of their grocery purchases, currently using Tesco Clubcard data.
Beyond gaining personal insight, users can choose to contribute their data to ongoing research. This initiative will enable the team to investigate how environmental perceptions align with actual behaviours at scale. Broader societal trends around geographic region, nutrition, and cost will also be explored to understand their links to consumer perceptions of the environmental impact of food.
By increasing our understanding of consumer beliefs and behaviours, we can more effectively design educational and behavioural interventions to promote lower impact food consumption.
Check out your own food environmental impact on the CDAT food web portal here.
