the system came about in part as a reaction to facial recognition. “My idea was, why don’t we simply consider the clothes someone’s wearing to understand demographics?” he said. “If I just showed you a shot of someone’s body you could probably tell me what gender they are. However, it turns out pointing a camera at someone’s chest or hips feels just as creepy as facial recognition. The idea was – what about people’s shoes?” The word “creepy” comes up a lot during discussions of in-store tracking. For retailers and data scientists, the aim is to find a way of obtaining information without coming across as intrusive. For Hoxton Analytics and the retailers using the technology, the answer is to look downwards. This tactic of avoiding the face and staring at shoes says a lot about how we, as physical beings, react to being watched. It suggests there are boundaries that do not exist on the internet. Set sights on our torsos and we feel invaded. But is making calculated judgments about a person based on their footwear actually any less invasive? For McCormack, the argument hinges on the fact that personally identifiable information isn’t being collected. “Right now, shops are doing lots of incredibly invasive things but we just don’t know about it. The angle Hoxton Analytics is taking on that whole thing is, well, if you know someone’s a male or a