Glossary
Service level indicator (SLI)
The specific, measured metric, such as success rate or latency, used to judge a service's performance against a target.
Also called: SLI
A service level indicator (SLI) is the specific, measured metric used to judge a service's performance, such as the percentage of requests that return successfully, the percentage completing under a latency threshold, or system uptime. It is a quantitative measurement, distinct from the target set against it.
An SLI differs from a service level objective in that the SLI is the observed number and the SLO is the goal that number is compared against; the same SLI, say, request latency, can support several different SLOs for different services or customer tiers. SLIs are typically sourced from application performance monitoring tools, distributed tracing systems, or infrastructure metrics, and the choice of SLI matters as much as the target: measuring average latency instead of a high percentile can hide the slow requests that frustrate users most.
SLIs matter because they anchor reliability conversations in a shared, verifiable measurement instead of anecdote, and a well-chosen SLI reflects what users actually experience rather than what is easiest to instrument. A common pitfall is selecting an SLI that is easy to measure but only loosely correlated with real user experience, such as server-side success rate that ignores client-side failures.
Last reviewed September 22, 2026