Explainer: How Local Data Startups Are Building Analytics for the Real World
There’s been a lot of excitement around in-store analytics recently as venture funding has begun to pour into the space. But the trend is one part of a much larger technological effort to quantify consumer activity in the real-world, bringing the same measurement about where we go, when we go there, and where we came from that Google Analytics and Comscore brought to the web years ago.
There’s a range of technologies at play in the market today, each of which has a unique set of risks, rewards, and capabilities for the business and consumers alike. What these technologies can provide is governed by a number of factors including their data source and collections, the service’s relationship with the consumers that they are measuring (passive, opt-in), and breadth or depth of that data, that vary in their use and intrusiveness.
Today, we can break down the effort into two buckets: technologies that can tell you what’s happening within a store, and those that help you understand broader activities (say, movements between locations). Each will play an important role in measuring our behavior in the real-world, creating a user layer, so to speak, to complement the business information data that companies in the local data ecosystem have developed over the past decade.
In the Store
There are two technologies that startups in the market are using to track and measure shopper behavior in-store. The first, used by companies like Euclid and RetailNext, analyzes the data generated when your smartphone interacts with a wifi router to provide Google Analytics-like statistics on the foot traffic into and around a store. Thanks to other attributes passed along, like the phone’s unique MAC address and the strength of signal, these services can also glean the amount of returning customers and the number who simply pass by the window but don’t come in. But can these techniques “track” users in the sense of following them throughout store?
Today, the answer is no. But the second technique, pioneered by companies like Prism Skylabs can. The technology uses the existing video feed streaming from a company’s surveillance camera’s to map the way consumers move through the store and provide retailers with remote access to these feeds in real-time.
The two technologies overlap, but the big difference here is that the video analytics approach offers businesses the ability to understand shopper behavior within the store itself. However, since a tool like Prism’s is not collecting identifiable information from shoppers like a MAC address, it cannot provide the type of information about the shoppers themselves (new, versus returning) that the wifi technology allows.
Out of the Store
With 145 million people in the U.S. owning smartphones, the mobile phone has become a revolutionary tool in quantifying the real-world. Consider Waze, the Israeli mapping startup that sold to Google for $1 billion earlier this year: the company has been able to to build one of the most complex roadway and transportation datsets at a fraction of the cost by tapping the collective data streaming from user’s devices.
As positioning algorithms improve, and the process of determining your location takes less battery life, developers will have the technical ability to track a user’s course throughout the entire day. The challenge for the industry is in finding ways to access that data in a scalable and privacy-friendly manner.
The Seattle-based startup Placed has developed a Nielsen-like approach to location analytics where the company incentives a panel of users to share detailed, but anonymized, location data from their mobile device with the company either by downloading a proprietary app or opting-in with an affiliate developer. With participants opted-in, the company can access extremely rich behavioral data for each user and start to understand where people are going, where they’ve gone in the past, and for how long. The company is monetizing the data through a number of products, including a recently released attribution service for mobile advertising.
Mobile ad networks have emerged as a valuable source of user’s location data, with a number of ad tech firms mining the billions of ad requests that come across exchanges everyday for insights into consumer behavior. Companies like PlaceIQ and JiWire have built products that use algorithms to index ad requests by a device’s ID, and then compare the location data that’s often included in these requests. They can tell you where a certain device has been in the past, as well as whether that user has appeared in an advertisers store after seeing an ad.
Given the number of new sensors that are entering the market every day, there are likely more innovations to come. Like the business location data ecosystem that’s emerged over the past decade, the companies building this new user layer of the local data stack will need to focus on merging and normalizing disparate information as much as mastering a single tactic.
Steven Jacobs is Street Fight’s deputy editor.