Using Foursquare Data to Predict Successful Store Locations

To see whether social media can offer any insights into how consumers respond to stores' locations, Anastasios Noulas at the University of Cambridge and colleagues analysed 35 million publicly available check-ins from 925,000 Foursquare users in New York over six months. They concentrated on information for three store chains: McDonalds, Starbucks and Dunkin' Donuts.

They used data about the popularity of 2/3 of the outlets to create and algorithm to predict that popularity (based on repeat visits, how far people would travel to the outlet, foot traffic, etc.) for the other 1/3 of the outlets. While each factor predicted some measure of popularity, the algorithm did quite well, "accurately ranking the top 10 per cent of locations 70 per cent of the time."

So, it's possible that social media data may become part of analysis used to find new outlet locations.

- New Scientist
- study from Cornell

Published Friday, August 16th, 2013

Written by Adena Schutzberg

Published in

Location Intelligence

© 2015 Directions Media. All Rights Reserved.