Granularity (Cost Granularity)

How finely you can slice cost data determines which questions it can answer.

Quick Definition

Cost granularity is the level of detail at which cloud spend is measured and reported, by hour, resource, tag, service, or business unit. Finer granularity enables more precise allocation and optimization, while coarse data hides the drivers behind a bill.

Granularity is the level of detail at which cost data is captured and analyzed: by month or by hour, by account or by individual resource, by service or by customer. Coarse data answers coarse questions. If you can only see monthly totals per account, you cannot say which deployment caused Tuesday's spike or what one customer costs to serve.

Finer granularity powers the practices that create value: anomaly detection needs hourly data to catch spikes quickly, unit economics needs per-customer or per-feature slices, and Kubernetes allocation needs container-level resolution because dozens of workloads share each billed machine.

Example. Two teams investigate the same 18 percent cost increase. One has daily service-level data and finds the cause, a misconfigured batch job, in twenty minutes. The other has monthly account totals and spends two weeks interviewing engineers.

More granularity means more data to store and process, so match depth to decisions rather than maximizing everywhere. Opslyft's Cost Visibility provides resource-level granularity, and the cost allocation guide explains how depth enables ownership.