Geodemographic Classification of Interactivity: Proof of Concept for Greater London

Geodemographic classifications are a proven tool for characterizing individuals according to the locations in which they live or work. With the advent of smart data, they are extensible to representing how social systems function, bringing focus to newly observed patterns of human interactivity; that is, the generalized patterns of observed movements that are revealed by user interactions with smart devices. This can greatly extend geodemographic classification beyond nighttime residence or workplace alone. Core to interactivity geodemographics is the use of dynamic sources of individual-level data such as Global Positioning System tagged in-app mobile phone impressions, which can be used alongside more conventional geodemographic indicators to develop a holistic representation of neighborhood connectivity and function. This article pioneers census output area (OA, or block)-level geodemographic classification of neighborhood interactivity, using disaggregate (point impression) in-app data for Greater London to demonstrate proof of concept. Census data are used to anchor interactions to conventional geodemographic domains. We demonstrate that in-app mobile phone location data can strengthen geodemographic representation of place alongside conventional measures of social similarity. Finally, we illustrate how geodemographic classification of interactivity can be deployed to better understand London’s nighttime economy and reflect on the generalizability of our findings in the smart data era.