What Closed Circuit TV Can Tell Retailers About Customers
For years, security camera footage went largely unused, kept in the dark in case of an extraordinary event like a robbery. But more and more, retailers are using it to better understand you on a day-to-day basis, using new technology to extract valuable information about the goings-on within a store.
Prism Skylabs, a San Francisco-based startup, has built software that analyzes video streams, measuring everything from overall store traffic to heat maps illustrating the way in which customers move through a store. The company, which closed $15 million round in October, also provides retailers with remote access to surveillance feeds, allowing businesses to replace some of the costly quality assurance testing, in which most larger retailers invest.
Street Fight caught up with Steve Russell, the company’s founder, to talk about the growing field of in-store analytics, the challenges in quantifying the real-world, and why the wi-fi tracking technology used by competitors falls short.
A big part of the company’s product is providing retailers with remote access to their surveillance camera feeds. What’s the challenge here?
What’s important to a retailer is having a high-quality image of the store and having one that is so privacy-protected that they can access it remotely. Previously, best practices would require the retailer to have to go to the store to do that. For instance, retailers spend a lot of money in terms of auditing or reviewing store behavior: you know ‘secret shopping’ their stores to make sure they’re providing the absolute best customer experience. Many times simply having imagery of stores themselves enables them to avoid those costs and do that auditing remotely or electronically.
There’s a number of other technologies being used to measure in-store behavior. From a competitive perspective, why go with video?
Today, the state of the art in-store measurement is a door counter, which simply lets you know how many people walk in the front door, or wi-fi alternatives that essentially serve as a better door counter. Really only video gives you the means to count people throughout the store and to understand the paths they take in aggregate through that store. That data is much more useful than anything you could get off from other classes of wireless sensors because its more granular and relates to every single person and event that occurs within the store not just those that happen to have their wi-fi antennas on.
There’s been some concern in the media, and in government, that the wi-fi-based analytics technology used by a number of your competitors infringe on a consumers right to privacy. What’s your take on the privacy concerns?
I have a real problem with [the wi-fi technology]. I hope the privacy issues get sorted out, but to me it comes down to the fact that wi-fi is just another frequency. Everybody kind of gets why you couldn’t be indexing consumers by their ID using a cell tower: it’s actually, patently, illegal. I think we’re just in this zone where Bluetooth and wi-fi are, for better or worse, treated differently for now.
But Prism’s technology offers a similar service. Where does value creation end and intrusion begin?
We never track or identify people — period. We’re both a privacy protecting alternative to those kinds of counting technologies and we’re also a more accurate way to count. Since not everybody has a cell phone, and not everybody has them on and there’s a lot of technical issues with tracking people by their cell phones. Our view is that it’s wrong to track people by their cell phones and ask them to opt out. But it’s also that you don’t get very good data [from wi-fi analytics].
There seems to be a split in the market between companies that want to build a CRM product, and fold in other streams of data, and those that want to expand into other forms of real-world anlaytics analytics. Where does Prism fit on that spectrum?
We’re taking a different approach than some of our competitors who are maybe seeking to be the universal dashboard for retailers. Our customers are among the largest and most advanced retailers out there and our strategy with them is to simply provide them an open platform and a new data source. Any company that today is collecting their sale data, their counting data, their other data and looking to understand their customers better can just add a data screen to their existing process rather than asking them to swap out the investments they’ve made for our platform.
Most of the adoption for in-store analytics has been among brands. What’s the opportunity further downstream?
There’s a sort of a glass-half-full application that [smaller] retailers can benefit from. A number of companies in analytics space have been forced to try to aggregate a lot of data in order to find some insight that’s sort of worthwhile for the business and that comes out of being expensive, tough to integrate. There’s a high threshold for what they need to prove to have value.
In our case, we give every retailer a fundamental value from the start: remote access into their businesses and a set of core metrics that they didn’t have before and as you get down to that and we do it in a way that deploys that low cost against their infrastructure. So if you put those three together I think there’s a more scalable business opportunity where you’re not fighting to do as much last mile integration to prove insight because the value flows more naturally out of the product.
Steven Jacobs is Street Fight’s deputy editor.