Currently for any Atlas instance, cost granularity is limited to the cluster. It would be a significant improvement to enable cost utilization tracking by database for storage and compute. Specifically, storage and cost records should be granular in time and label, where labels can be assigned to databases. That is, as a common example with GCP for example, costs should be tracked by second, and in the billing data, aggregated to the hour by label.

This will allow chargeback, showback, and improved allocation methods for modern FinOps practices. As it is, MongoDB Atlas falls well short of any ability to align costs in a manner that many cloud customers expect and are using from all modern cloud providers.

I would be happy to discuss this with your billing development product manager or other leads, and will appreciate your consideration of this improvement.