The lie big data tells us.

The exponential rise in the use of smartphones has created a mountain of mobile location data that essentially tracks the geographic position of users. This data has helped usher retail location intelligence into the era of “big data”. There are two general ways it’s used. First, to better understand customer behavior and then use it to effectively tailor and target marketing efforts to this behavior.

Second, it’s used in retail site selection and other capital expenditure decisions. The process involves geofencing an existing store or a radius around a new potential site (i.e. 100 meters) and then collecting the mobile data on where the visitors are located overnight (or during the workday for some applications). The data is visualized on a map and then used to define a store’s trade area or potential store’s trade area. Trade areas developed using mobile phone data are often seen as “data driven” and therefore considered more accurate than theoretical circles or drive-times, both of which require a size assumption (i.e. distance or time).

While mobile data may improve the overall accuracy of trade area boundaries, it doesn’t solve the major problem with it – double counting consumers. The same consumers in your trade area are also shopping in nearby competing trade areas, each of which count many or all of the same consumers to be in their “trade area”. So, like theoretical circles and drive-time trade areas, mobile phone data also grossly overestimates the size of a trade area, but now with a deeper level of confidence that it’s accurate. This is the inadvertent lie big data tells retailers – that any consumer who enters your trade area will purchase their full share of wallet in your trade area. It’s called spatial equilibrium in the language of economics and it rarely occurs within one trade area. In real life, there are always consumer surpluses and leakages between trade areas.

It’s a problem mostly ignored in retail location intelligence, and yet its primarily responsible for the fact that nearly half of all Canadian retail is located in crowded low revenue cluster traps where there are too many retailers for the size of the market. It means that larger chain retailers seldom enjoy a location advantage but instead stumble along in mediocre networks trying to stimulate and attract customers by offering yet another sale. In the end, retailers wrongly assume they’ve reached the capacity of their brand.

Big data requires big context. This means that trade area definitions using mobile phone data must include the impact of retailers in nearby competing shopping centres for them to tell the complete story. Future advancements in location intelligence hinge critically on modelling trade areas as a system that better explains the complexity and mobility of consumers in the retail marketplace.

See our Free Network Report for a free evaluation of the capacity of your brand in Canada.