Software teams have a name for it: the bus factor. It is the number of people who could get hit by a bus tomorrow before the project grinds to a halt. In a lot of small businesses, the honest answer is one. You.
But the owner is rarely the only single point of failure. There is usually a bookkeeper who is the only one who knows how the books actually reconcile, a lead technician who is the only one who can handle the hardest jobs, an office manager who is the only one who knows which vendor to call when something breaks. Each of these people is doing good work. Each of them is also, quietly, a risk you have never priced.
What a Single Point of Failure Actually Costs You
A single point of failure is not just about someone quitting. It shows up in smaller, more frequent ways: the person who is the only one who knows the process takes a real vacation and everything slows down. They get sick during your busiest week and there is no one to cover. They ask for a raise and you cannot say no, because you both know what happens if they walk.
That last one is the cost that never shows up on a spreadsheet. When one person holds irreplaceable knowledge, they hold leverage over you whether they intend to or not. Compensation conversations stop being about market value and start being about what it would cost you if they left tomorrow. That is not a healthy foundation for either side of the relationship.
How to Find Every Single Point of Failure
Most owners think they know where their risk sits. Most are wrong, because the risk is usually hiding in a function they stopped paying close attention to once it was running smoothly. Here is a simple audit to surface it.
1. List Every Critical Function
Write down every function that, if it stopped working for a week, would meaningfully hurt the business. Payroll, client delivery, key vendor relationships, invoicing, quality control, whatever applies to your model. Do not stop at the obvious ones.
2. Ask the One-Person Question
For each function, ask: how many people in this business could step in and run this competently with two weeks notice? If the answer is one, you have found a single point of failure. If the answer is zero, because it is you, you have found the biggest one in the company.
3. Rank by Impact, Not Discomfort
Owners tend to fix the single points of failure that make them personally uncomfortable first, not the ones that would actually hurt the most if they failed. Rank your list by what would genuinely cost you money, clients, or reputation if that person disappeared tomorrow. Fix in that order.
Three Ways to Close the Gap
Cross-Train Before You Need To
The most direct fix is also the one most owners avoid, because it takes time away from the work that feels urgent. Pair your single point of failure with a second person for real, hands-on training, not a one-time shadow day. Build it into a quarter, not a weekend. The goal is a second person who could run the function at seventy percent competence tomorrow, not a perfect backup.
Document the Judgment, Not Just the Steps
Most critical knowledge is not the steps of a task, it is the judgment calls: which client gets extra flexibility, which vendor to trust when two quotes look similar, what "good enough" looks like versus what needs to be redone. Capture those decisions in writing as they happen, not from memory months later. This is the difference between a checklist that covers the routine day and documentation that covers the hard one.
Build Redundant Relationships, Not Just Redundant Skills
If your lead salesperson is the only one who has a real relationship with your three biggest clients, that is a single point of failure even if someone else could technically do the job. Get a second person into those relationships deliberately, on calls, in meetings, cc'd on the threads that matter, well before you need them to take over.
The Test That Tells You If You Are Done
You will never eliminate every single point of failure completely, and you do not need to. The test is not perfection, it is whether the business can absorb a surprise. Pick your riskiest function and run a real test: have the backup person handle it for a full week while the primary person is genuinely unavailable, not just on call in case something goes wrong.
If the week goes fine, you have converted a risk into a system. If it does not, you just found the gap while it was still cheap to fix, instead of finding it during an actual emergency.
Start With the Scariest Name on Your List
Every business has a name that comes to mind immediately when you ask who could not be replaced right now. Start there. Not because it is easy, but because it is the single point of failure most likely to actually fail you, and the one that will cost you the most in money, time, or leverage over your own business if you wait.
A business that depends entirely on irreplaceable people is not a business. It is a collection of personal relationships you happen to be paying for. Fixing that, one function at a time, is what makes the difference between a company you own and a job you cannot leave.
Build the Full System
Built to Run includes the complete framework for auditing operational risk, cross-training your team, and building a business that survives the loss of any single person, including you.
Get Your Copy →