Exploring consumer behaviour in dining contexts
Introduction
In the rapidly evolving landscape of consumer research, understanding behaviour in dining contexts has become increasingly vital. This need has paved the way for innovative tools that can capture real-time interactions in social settings. Enter the AI Witness Tool, an advanced setup featuring a lamp camera equipped with audio recording capabilities. This tool promises to revolutionise how we measure consumer behaviour during dining experiences by observing interactions with food and with each other.

Measuring consumer behaviour in social dining contexts
The AI Witness Tool is strategically placed on the table during dining experiences to record subtle interactions among diners and with food. By linking audio recordings to video footage, the tool allows for implicit measurement of consumer behaviour, providing valuable insights into how people engage with their meals and one another in a social setting.
Pilot study overview
A pilot study was conducted to assess the effectiveness of the AI Witness Tool in recording these interactions and its ability to be trained to recognise and monitor specific objects during these exchanges.
What was done?
The pilot study aimed to evaluate the AI Witness Tool’s capabilities by examining its performance in a controlled environment. Participants were observed to determine if the tool could successfully record interactions and recognise specific objects when an interaction occurred.
Where was it conducted?
The study took place in an immersive room designed to replicate a restaurant setting, providing a realistic backdrop for evaluating dining behaviours.
Who participated?
The study involved eight individuals, each entering the immersive room individually and accompanied by the same person. This setup facilitated focused observations of interactions between the two individuals.
Study setup
During the study, two participants sat at a table facing each other, with a selection of specific food products placed in front of them:
- Carrots
- Cherry snack tomatoes
- Snack cucumbers
- Hummus
- Water

Objectives of the pilot study
The primary objective of this pilot study was to determine whether the AI Witness Tool could effectively record interactions and be trained to recognise specific objects during these interactions. By evaluating the tool’s functionality, the researchers aimed to lay the groundwork for more extensive studies on consumer behaviour in dining contexts.


Next steps
The pilot study’s findings will inform several next steps in the development and application of the AI Witness Tool:
- Testing video quality: Future studies will assess the quality of video recordings in various settings, comparing conditions such as daylight and nightlight to determine optimal usage environments.
- Enhancing analysis: The aim is to improve the analysis capabilities of the AI Witness Tool by developing features that can recognise different individuals present in the video, providing insights into group dynamics.