Error Budget
An error budget is the amount of unreliability allowed by a service-level objective.
What is Error Budget?
An error budget is the acceptable amount of failure implied by a service-level objective over its measurement window. If a request-success SLO is 99.9%, the remaining 0.1% is the budget. Teams use budget consumption to balance reliability improvements with feature delivery, supported by an explicit operating policy.
Example in SaaS
A SaaS expects one million valid requests during a window and targets 99.9% success. The illustrative budget is 1,000 unsuccessful requests. If the service consumes most of it early, the team may prioritize a recurring failure before expanding a rollout. A time-based availability objective would use time rather than requests.
What to watch for
An error budget is not permission to deliberately break the service or ignore affected customers. The policy should specify action when failures accumulate quickly and account for different customer journeys. Do not combine a request-based budget with downtime minutes without changing the definition. A zero-failure target leaves no budget and can impose costs that need to be justified by customer needs.
Source and further reading
Related terms: Service-level objective, Feature flags.