Another experiment: using marked areas on Google Maps for various purposes in e-commerce: finding the optimal warehouse, searching for accessible stores for self-pickup or the best delivery service, and maybe even the very fact of being able to sell a product or service to a customer from this zone.
It works like this: the customer enters an address, and the system identifies it as one or more major zones. Different components of the system depend on these major zones, not on the minor components of the address such as the postal code.
At the same time, I got to grips with developing on Google AppEngine. The issue is that determining the polygon (zone) that includes a point on the map (where the customer is), for the situation of “many zones of complex shapes”, can potentially be a quite “heavy” computational task. If there’s a possibility, it’s better to handle it immediately on a cluster that can easily scale, and even better, do so autonomously. And this case is excellent for Google AppEngine, where Google DataStore is used to store polygon parameters, and Google Memcache for storing cache.
https://hybrismart.com/2016/10/19/geofencing-in-hybris-custom-shipping-zones/
Hi, I'm Rauf Aliev. 