If there’s one thing that retailers want, other than the sound of ringing tills, it’s information about their customers: Where do they come from? What do they do? And how can we better encourage them to exchange cash for goods?
This has spawned all manner of labour-intensive efforts over the years, from customer surveys to cashiers who earnestly solicit your postal code at the Old Navy checkout to Air Miles or Optimum customer loyalty programs that diligently catalogue your purchases.
But what if there was a way to generate customer profiles without interacting with the customers at all?
RetailGenius, a product from a Toronto startup called Viasense, promises to algorithmically generate customer profiles based on a remarkable source of data: Anonymous location data that’s collected by big mobile carriers, from the passive pings that every single cellphone sends out as it goes through the day.
The data that RetailGenius uses is anonymized – it doesn’t have any way of knowing whose cellphone belongs to who; it simply has a gigantic plot of where thousands of cellphones were at any given time.
“We create a unique identifier between those signals, and we can see those signals move throughout the city,” says Mossab Basir, RetailGenius’ founder. “We can see those changes in your location but we never really know who it is.”
What the product does next is intriguing: Based on some 50 million pieces of location data a day, RetailGenius crunches the numbers to make inferences from where each cellphone spends its time, and generates customer profiles by the thousands.
For instance, if a given cellphone spends the hours between 7 p.m. and 6 a.m. in a single area, it’s a good bet that its owner lives there. If that cellphone spends its working hours downtown five days a week, its owner is probably a daily commuter. And if it visits a given retail store once a week, a picture of its owner’s habits living and shopping habits starts to emerge.
By lumping these inferred profiles together, RetailGenius can give retailers a picture of who walks through their doors. For instance: What are the top 50 postal codes that are represented in their customers? What kind of volumes of customers are arriving at the store? How long do they stay?
Of course, the key to this product is getting the data in the first place. Every cellphone stays in touch with the cell towers around it, even when you’re not using it – that’s how it always has a signal ready to use when you wake it up. These towers are arranged in a grid pattern that resembles cells (thus, cell-phones) and they’re always paying attention to how strong your signal strength is.
When your phone starts leaving the range of one cell tower, your call (or data) gets cleverly handed-off to the next-best tower. By triangulating these signals and factoring in signal strength, cellphone carriers can get a reasonable idea of where your phone is, even without features like GPS.
Basir says Viasense gets its data from social networks as well as major carriers, who are cautiously entering the “bulk data” marketplace, selling anonymized data for analysis. You’ll be hard-pressed to find reference to this on cell carrier websites; however, a Rogers spokesperson confirmed that the company has “a small number of customers who use aggregate, anonymous locator information to predict traffic patterns.”
Basir, who started his career in corporate branding before moving into startups and technology-focused marketing, says Viasense has raised a half-million dollars in angel funding. The 50 million location events a day his firm is processing – covering much of the GTA, at present – is just the start, he says. “The amount of data that’s out there from a big data perspective – that’s what really excites us.”