Identifying Single Points of Failure

Identifying single points of failure means finding the components, roles, information or suppliers whose loss could break a service because no viable alternative exists. The analysis must look beyond hardware to people, knowledge and suppliers, and prioritise by impact on outcomes.

• Map service dependencies. Start by understanding what the service relies on, with clear decision rights and escalation paths where several functions or suppliers contribute.
• Test redundancy and failover. Assuming redundancy has been tested is a named pitfall. Failover test success is one of the measures that helps, alongside the number of unmitigated single points of failure and overall recovery capability.
• Include people and knowledge. Looking only at hardware misses important risks. In the article's example, a resilient platform still has a single expert authorised to recover a critical integration, creating a people-based failure point. Ignoring sole experts and unique suppliers is a common mistake.
• Prioritise by outcome impact. Focus effort where a failure would most harm the consumer outcome. After the first cycle, ask whether recovery capability improved for the intended reason and whether any group experienced a worse result, and review findings with delivery teams and affected consumers. Evidence of a risk should lead to an owner and a proportionate action, not just another note in a report.