Split shared costs by usage metrics

Use metrics from observability platforms like Datadog or Prometheus as allocation drivers so shared costs follow actual consumption.

Shared infrastructure costs like network transit, RDS, and S3 are notoriously hard to allocate fairly. Even-split or pro-rata rules are easy to set up, but they rarely reflect who's actually driving the spend. When one service accounts for most of the traffic but only pays a fraction of the bill, your cost model is telling the wrong story.

Cloud Intelligence™ now supports a driver-metric split mode that allocates shared costs based on external usage metrics, starting with Datadog.

Datadog metric exports: Define named metric queries using full Datadog query syntax, including tag group-bys. A Datadog integration sync pulls a trailing three-day window into Cloud Intelligence™, keeping your usage data current alongside your billing data.

Driver-metric split mode: Select "driver metric" as the distribution method when splitting costs on a report. Each target's share is proportional to its share of the chosen metric per time bucket, so allocations track real consumption instead of static rules.

No silent defaults: When a time bucket has no driver data, the origin cost stays unsplit rather than being silently spread evenly.

media

To get started, set up your metric exports under Integrations → Datadog, then apply driver-metric splits on any Cloud Analytics report.

Available to customers with a configured Datadog integration.