A single point of failure is any part of a business process where one person, one system, or one document is the only thing standing between normal operations and a significant problem. Every business has them. Most businesses do not know where theirs are until one of them fails.
Why single points of failure are hard to see ¶
Single points of failure are often invisible precisely because they are working. The person who is the only one who knows how to run a particular process is doing it reliably, so nobody notices the risk. The spreadsheet that lives on one person's desktop is being updated correctly, so nobody asks whether there is a backup. The problem only becomes visible when the person is ill, or leaves, or the laptop is stolen. By that point, the operational impact is immediate and the time to fix it is gone.
How to find them ¶
The most reliable way to find single points of failure is to ask a simple question about each critical process: what would happen if the person who normally does this were unavailable for three weeks? If the answer is 'we would struggle' or 'we are not sure', that is a single point of failure. A process review conducted by Hilt Core Vault systematically asks this question for every process we map. In a typical review of a business with forty to eighty staff, we find between four and nine significant single points of failure.
Fixing them without over-engineering ¶
The fix for a single point of failure does not need to be complicated. In most cases it involves three things: documenting the process so that someone else could follow it, identifying a named backup person who is trained on the process, and scheduling a regular check to confirm the documentation is still accurate. The goal is not to eliminate all risk. That is not achievable. The goal is to ensure that no single absence or failure can bring a critical process to a halt without warning.
Prioritising which ones to fix first ¶
Not all single points of failure carry the same risk. The ones to fix first are those where the impact of failure is highest and the likelihood of failure is not negligible. A process that is critical to revenue or to regulatory compliance and that depends on one person who has been in the role for twenty years and is approaching retirement is a higher priority than a process that is mildly inconvenient to lose and that has three people who could cover it. Prioritisation is a judgement call, but it should be a documented one.
If you want to know where your business's single points of failure are, a process review is the most direct way to find out. Get in touch to talk through what that would involve.