← Back to all stories


Measuring mobility for major forecasting gains

phData, a full-service machine learning and data analytics consulting company, uses Mapbox Movement Data to drive major improvements in the accuracy of forecasts they create for clients in logistics and retail. For many of these companies, having an accurate forecast can make or break whether shipments will make it on time.

“phData has generated forecasts that are up to 20% more accurate when using Mapbox Movement as an input. We tested this by forecasting shipments for a logistics company during 2020. The mobility forecast accuracy in one instance was near 80% for a given week. Depending on the size of the company and the complexity of its logistic operations, a 1% increase in planning accuracy could result in millions of dollars saved.”

Elizabeth Dinevski, Director of Data Science, phData

phData helps its clients' analytics teams by educating people, optimizing technology, developing world-class data products, and providing sustainable results. By partnering with clients to change the way they interact with data and analytics tools, phData transforms them into outcome-driven organizations. This allows its clients to focus more on understanding their business and less on the process of collecting and analyzing forecasting data.

phData found that human behavior is the best indicator for demand and they use Mapbox Movement, the world’s largest privacy-forward mobility data set built from 700 million monthly active users and updated daily, to serve as a measure for human behavior. The more mobility in a given area, the more economic activity we can expect to see, and this can be used as an input for forecasting to drive better strategic business planning.