Is My Azure Bill Too High for Data Analytics?

In my experience, the overall $6k isn't too extreme for a B2B SaaS dealing with analytics, but you should certainly look into that 50% DTU cost. It's often the case that teams overestimate their needs. I'd suggest checking out vCore General Purpose or Hyperscale options if you have heavy read reports. Plus, analyzing your Cost Management breakdown can reveal where most of your expenses are coming from.

We realized that a significant portion of our bill came from Azure Data Factory too, so we moved a lot of that functionality back to .NET code and used Hangfire for scheduling. It really helped cut costs!

For Azure Data Factory, it's essential to check the Integration Runtime you're utilizing. Are you managing the autoscaling features properly? Regarding your SQL costs, switching from the DTU model to a vCore model can save you 25-40% since you’ll be paying based on actual cores used. Also, don't forget the cost impact of how frequently you're polling Salesforce; reducing that can save you money!

What tier are you currently using for your SQL setup? This can impact your overall costs significantly.

If your bill is around $6k, it might be worth looking into different SQL SKUs instead of continuing with the DTU model. Non-DTU SKUs can offer higher IOPS which might suit your workload while potentially allowing you to use a smaller instance. Also, consider implementing reservations and savings plans if you haven’t done so already!