Variations are normal, and can be caused by multiple factors:
- The creation or deletion of resources (e.g. creating an Azure Cosmos database, deleting a virtual machine).
- Increase or decrease in usage (e.g. number of requests, bandwidth, CPU, operations used)
- The scale up or down of services (e.g. app service going from S1 to S2)
Identifying variations are a great start when you are investigating sudden increase or decrease in spending.