It's important for customers to fully understand the pricing metric for each piece of technology being purchased. This is true especially with cloud vendors, who can have complex pricing models for services that are very new and otherwise hard to price. Since a lot of the technology services offered by cloud vendors are new or offered in a way that is new, the pricing mechanisms can be very different and require a thorough understanding of how much of the service will be used, so there are no surprises in the cost of usage. For base infrastructure services, such as virtual instances or blob storage, the metric for pricing is usually per hour or per month, with the rate going up as the size goes up. However, for managed cloud services, the price can be more complicated. The following are some examples from popular AWS services used in cloud native microservice architectures: AWS Lambda and Amazon DynamoDB.

Given the complexity and options for pricing, it's critical for the design team to understand how these services will be used in the architecture. Even small changes to the size of transactions, runtime of functions, or other aspects can cause a significant increase in the bill if the system scales high enough.