Catchment Area Analysis for a Café in Seattle using Azure Maps

 

Jump directly to the code, if you'd rather:

http://bit.ly/2lAC2jS

 

Catchment is defined as the sphere of influence from which retailers draw their customers. Azure Maps  can be used by Retailers to better understand a location and analyze business drivers and disruptors from a spatial perspective. By building isochrones around the store location retailers can search within the Isochrone to determine the demographic mix, parking spaces, competitors and other information that could potentially help drive traffic to, or from, the business. 

I have been experimenting with Jupyter Notebooks as a great way to create and share my Azure Maps experiments with live code, visualizations and narrative text. Jupyter Notebooks are also an excellent way to do ad-hoc analysis using its interactive code-base environment. 

 

The Route Range (Isochrone) API  helps calculate a set of locations that can be reached from the origin point based on fuel, energy, or time budget that is specified. A polygon boundary (or Isochrone) is returned in a counterclockwise orientation as well as the precise polygon center which was the result of the origin point. This API is excellent for determining the Catchment Area for a retail location. In my code experiment, I used a 10-minute walking zone around a Starbucks store in downtown Seattle. Once I built the Isochrone around the store, I could use the Search Geometry endpoint to try and find business drivers and disruptors for that Starbucks store.  

 

For instance, fastfood chains are now a big disruptor to coffee chains like Starbucks and could be considered business disruptors. I set the search term in my code to "fastfood" and it found 11 fastfood establishments within a 10 minute walk of the Starbucks location:

 

1 . Gyros Place
2 . Peach Labs
3 . McDonald's
4 . Hot Dog Joes
5 . Taqueria
6 . Gyro Stop
7 . Jack in the Box
8 . McDonald's
9 . Buffalo Wild Wings
10 . Street Hot Dog Stand
11 . Monster Dogs

 

On the other hand, an abundance of parking locations is a business driver. When I change the query to look for Parking within a 10 minute zone of the Starbucks, I do get plenty of options (it shows over 50 parking lots within the isochrone). Try experimenting by changing the query to different business drivers and disruptors. Establishing a score for the business drivers and disruptors will enable you to establish a profitability score for your location. 

 

You can download the code from my github repo. You can experiment with the downloaded code using Azure Notebooks. I will be posting a detailed blog shortly but the embedded narrative in the Jupyter Notebook, is, I hope, pretty detailed.