// function to pull cookie value Skip to main content


The Value of Trajectories

We are constantly ideating, innovating, and envisioning new ways to make our product better. Stops data is one of the main assets we use to create our products, which is why it’s fundamental to understand the algorithm that produces stops and how we can leverage the information we process along the way.

Stops data is obtained by algorithmically clustering data points. We collect these data points daily and compute stops into “points” based on where and how long a device has stopped. Beyond these datasets, it’s valuable to know what happens between two different stops to show how people move between two destinations. Understanding movement on this level creates the potential to identify travel routes, trade-area analysis, and traffic patterns.

This past year, we introduced a new, innovative dataset called trajectories. A trajectory is defined as the path traveled by a given user between two consecutive stops within a given day. Normally, our clustering algorithm generates stops and our POIs are used to convert stops into visits. But rather than singular points or a straight line between two consecutive stops, a trajectory maps the actual path traveled from point A to point B.

What Are Trajectories and Why They Are Valuable

These datasets are immediately available in the Spectus cleanroom and can be used to easily perform origin-destination analysis. With trajectory data, origin-destination analysis allows you to map the route, speed, and frequency of travel between points for a defined period of time. This results in a multitude of interesting and useful applications, like:

  • Analyze trajectories starting or ending within a given location to easily represent a catchment area of a given store or look at travel patterns to/from a given region
  • Isolate users passing by a specific point of interest such as a billboard, then understand where they went before and after
  • Leverage valuable metadata to filter or enrich your analyses (i.e. maximum speed reached in a given trajectory, length of a trajectory, and duration of travel)
  • Understand urban traffic patterns over time for transportation analytics use cases
  • Employ trajectories data for governmental organizations to analyze who is visiting a city, from where, and their mode of transportation
  • Show trade-area at the street level by analyzing which routes and streets customers take to get to a store
Street level trade-area data for two Walmart stores in Phoenix
Street-level trade area data for two Walmart stores in Phoenix

Interactive Web App to Demonstrate Trajectories Use Case

We have already been able to demonstrate the value and validity of this data in several different use cases. Our team analyzed trajectories from a single day in Manhattan and created an interactive web app to allow easy exploration of this feature. The circles represent interesting hot spots, like major airports, Grand Central Terminal, and the Statue of Liberty. 

Click on these circles or around the Manhattan area to see the trajectories to and from that point.

Note: This analysis was made privacy-safe by our PEM methodology and trajectories starting or ending at inferred home locations have been removed. We also removed trajectories where there were too few devices in the starting and ending points.

Interested in learning more about Trajectories and how they can add value to your business?

Let's talk