Evaluate the cost of distributing data globally
When evaluating the cost of a solution that replicates across multiple regions, you should calculate the cost as a multiple of the single-region cost relative to the number of replicas. There are two ways to calculate this cost based upon the type of throughput being used, standard provisioned throughput or autoscale.
Set throughput to standard provisioned
For standard provisioned throughput, the core formula is RU/s x # of regions. For example, consider a solution that uses approximately 1,000 RU/s per hour; data is written to one Azure region and replicated to five more regions. The formula for this throughput would be:
1,000 x (1+5) = 6,000
The account would be billed for 6,000 RU/s used at a per-hour rate.
Set throughput to dynamic autoscale
For dynamic autoscale throughput, the core formula is Minimum RU/s x # of regions + RU consumption per region. For example, consider a solution that has autoscale of 1,000 RU/s. The minimum RU/s for this throughput is 100 RU/s. The cost for that container in every region when it isn't in use is 100 RU/s. When the container in that region needs to scale, the cost is whatever the container scales to for every hour it's scaled out. The formula for this throughput would be:
100 x (1+5) = 600 + Consumed RU per hour, per region