Observability has quietly become one of the fastest-growing cloud cost centers, rivaling compute as teams over-ingest, over-retain, and over-query telemetry with little financial accountability. In mature environments, observability can consume up to 30% of total cloud spend, often without a clear connection to business value.
This session applies FinOps Framework principles to observability, showing how to implement granular show back by workload and service, control ingestion through dynamic sampling, apply tiered retention policies, and optimize high-cost queries, delivering 40%+ cost reduction while maintaining SLOs. Using real-world architecture patterns and cost breakdowns, we will challenge common myths such as “high cardinality is inevitable” and illustrates proven techniques like index pruning, decentralized processing, and query discipline. Finally, we will look at chargeback and accountability models that drive better engineering behavior and help organizations progress from Crawl to Run maturity, turning observability from an uncontrolled tax into a governed, value-aligned capability.