SafeGraph’s newest dataset Spend contains anonymized debit and credit card transaction data aggregated to individual places in the U.S. and is released on a monthly cadence.
This data complements Patterns, which contains aggregated raw counts of visits to POI from a panel of mobile devices, answering how often people visit, how long they stay, where they came from, where else they go etc.
Since many of our community members work with Patterns data, I wanted to share this tutorial I’ve been working on correlating Spend and Patterns. Here’s a link to the notebook.
In this notebook, I used Spend at Chipotle Restaurants and Patterns foot traffic at nearby gyms. Gyms fall into the category of having billing that might not take place every time a customer interacts with the business. We will define a Chipotle and a Gym as being nearby if they fall into the same Placekey hexagon. The notebook aims to model how delta in transactions at Chipotle can be observed in the foot traffic at neighboring gyms.