The single point of failure every contractor is carrying

Ask a contractor's CEO which person they could least afford to lose and they answer in under five seconds. It is usually a project controls lead, a commercial manager, an estimator or a planner. Sometimes it is the person who "knows the system". The speed of the answer is the finding. Everyone in the business already knows where the single point of failure is. Nobody has written it down, priced it, or built a plan for the week it goes.

What a single point of failure looks like in a project business

It is rarely one thing. In the audits we run it tends to be a chain. One person owns job cost. That person's method lives in a workbook only they understand. The workbook pulls from a legacy system the vendor stopped supporting three years ago, a platform running on borrowed time that nobody has had the appetite to replace. The report leadership steers on comes out of that workbook on a Friday, so by the time the number is read it is a week old. Each link is survivable on its own. Together they mean the business is running on the assumption that one person never gets sick, never resigns and never makes a mistake in the wrong cell.

The same pattern shows up in commitments and change. Change orders tracked by email. Client commitments held in someone's head. A licence register that is really a spreadsheet that one coordinator keeps current in their own time. None of it is a system. All of it is exposure.

What it costs, before it fails

Leaders tend to price this risk at zero until the day it triggers, because it does not appear on a P&L line. It is there anyway. Decisions made on stale numbers. Disputes that cannot be evidenced because the trail is in an inbox. Bids lost to competitors whose controls are tighter and whose margins are therefore sharper. Rework nobody measures. And an unpriced dependency on a handful of people, which your best people are quietly aware of and which shapes what they ask for at review time.

Then there is the day it actually fails. A resignation in the wrong week and a missed contractual milestone. A crash in the unsupported system and a client asking why the business could not produce basic operational data for a week. A claim that needs three years of change history you cannot produce, because it was never governed. Industry data says 45% of construction firms already report project delays because of workforce shortages, and that is before you count what happens when your one indispensable person walks.

Why it does not get fixed

Because the person carrying it is competent, and competence hides risk. The workbook works. The report lands. The client is happy. Fixing it looks like a project with no visible payoff and a lot of visible disruption, so it stays on the list behind whatever is on fire this month. That is deferred investment, quietly compounding. Every year it is left, the eventual fix gets more expensive and more disruptive than the one that was avoided, and it usually gets fixed only after the failure, at the failure's price, on the failure's timing, by a business explaining to a client or a board why it happened at all.

The order to remove them in

Do not start with the software. Start with an audit that names the specific failure modes, with evidence: who holds what, in which system, with what fallback, at what cost if it stops. Then agree one priority risk with your operations lead in the room, not a list of twelve. Remove that one with a fixed-scope build that mirrors how the work actually runs, validated side by side against the outputs you already trust, and signed off by both sides before it touches production. Then keep watching it, because a fix nobody monitors is a new single point of failure, and a business carrying a prior failure that was never fully rectified cannot afford to let the same exposure resurface a second time.

That order matters. Findings first means you never buy a build you do not need. One risk at a time means the operation keeps running while it is fixed. Validation before production means the people who used to carry the risk trust what replaced it, which is the difference between a system that is used and a system that is worked around.

If you answered the five-second question with a name, that name is your audit's first line. The rest of the report is about how much it has been costing you to leave it there, and whether growth has already outrun the controls built to catch it.