Case study
International Tourist Model
Summary
Third party mobile geolocation data can accurately track where domestic visitors live and shopping behaviours, however low sample rates mean it is less reliable in understanding the movement of international visitors.
The International Tourist Model has been developed to address the inherent limitations of mobile geolocation data in accurately capturing the proportion of international tourist visits to specific locations. By integrating VisitBritain data with CACI’s Local Footprint dataset, this model offers a comprehensive and robust solution.
Industry
Technology
Products used
Challenge
Mobile app data, which relies on location “pings” from smartphones to analyse customer behaviour and footfall patterns, is a powerful and highly accurate tool for identifying domestic visitors. However, due to lower adoption of UK-based apps among international tourists, the use of VPNs, and varying data protection regulations across countries, mobile app data struggles to reliably identify international visitors.
To overcome this, CACI has developed the International Tourist Model with three core objectives:
- Accurately represent the proportion of international tourist footfall at different locations.
- Provide insights into the continental origins of international tourists.
- Report these data points across different time periods.
Third-party mobile geolocation data is less reliable for tracking international visitors due to low sample rates.
Mobile app data struggles to identify international visitors because of lower app adoption, VPN usage, and varying data protection regulations.
The International Tourist Model integrates third party data with CACI’s Local Footprint dataset to accurately represent international tourist footfall.
Solution
By leveraging government-published data on inbound and domestic tourism and blending with CACI’s Local Footprint data, and third-party geolocation data enabled us to infer the relative presence of international tourists through a data cleansing and modelling process.
Results
The model has demonstrated clear value and has already been implemented across multiple projects, delivering tangible benefits to clients to allow these to understand the true mix of user groups interacting with their assets.
