Observability challenge #1: Fragmentation and complexity
Traditionally, organizations have deployed multiple observability tools across their technology stacks to address distinct needs like monitoring logs, metrics, or traces. While these specialized tools excel individually, they rarely communicate well, resulting in data silos. This fragmentation prevents teams from gaining comprehensive insights, forcing devops and SRE (site reliability engineering) teams to rely on manual integrations to piece together a full picture of system health. The outcome is delayed insights and an extended mean time to resolution (MTTR), slowing down effective issue response.
Additionally, organizations now need to incorporate data streams beyond the traditional MELT (metrics, events, logs, and traces) framework, such as digital experience monitoring (DEM) and continuous profiling, to achieve comprehensive observability. DEM and its subset, real user monitoring (RUM), offer valuable insights into user interactions, while continuous profiling pinpoints low-performing code. Without integrating these data streams, teams struggle to link customers’ real experiences with specific code-level issues, resulting in data gaps, delayed issue detection, and dissatisfied customers.
Observability challenge #2: Escalating costs
The cost of observability has surged alongside the fragmentation of tools and the growing volume of data. SaaS-based observability solutions, which manage data ingestion, storage, and analysis for their customers, have become particularly expensive, with costs quickly accumulating. According to a recent IDC report, nearly 40% of large enterprises view high ownership costs as a major concern with observability tools, with the median annual spend by large organizations (10,000+ employees) on AIops and observability tools reaching $1.4 million.