Where Contingency Goes Wrong: Sizing Reserves Without Fooling Yourself
Contingency is the part of a project budget that exists because the future is uncertain. Done well, it is a disciplined estimate of what known risks could cost. Done badly, it is a round number someone added at the end to feel safe, then spent in the first month on things that were never risks at all. The disruption of the past year exposed a lot of weak reserves: schedules and budgets with no give in them buckled the moment a shipment slipped or a key person was off sick. Here are the mistakes that show up again and again, and how to avoid them.
Confusing the two kinds of reserve
Contingency reserve and management reserve are not the same thing, and treating them as one pot causes trouble. Contingency reserve covers identified risks — the things on your risk register that you can name and estimate. It sits inside the cost baseline and the project manager controls it. Management reserve covers the unknown unknowns — work that was genuinely unforeseeable. It sits outside the baseline and usually needs sponsor approval to release. When teams blur the two, they either raid the risk money for scope they simply forgot to plan, or they freeze reserve they were entitled to use.
The mistakes that drain reserves quietly
Picking a flat percentage and stopping there. Ten percent across the board ignores that risk is unevenly distributed. A novel, first-of-its-kind work package may need thirty; a repeat job you have done ten times needs almost nothing.
Padding every line item. When each estimator hides a private buffer, the total contingency is invisible and uncontrollable, and Parkinson's law guarantees it all gets spent. Hold reserve openly at the project level instead.
Sizing it once and never revisiting. Contingency should burn down as risks expire or occur. If you are halfway through with most risks behind you and the reserve untouched, some of it should be released, not quietly absorbed into new scope.
Treating reserve as a slush fund. Spending contingency on a change in scope, rather than on a realized risk, hides the fact that the project grew. Every draw should map to a specific risk or change request.
Ignoring the schedule side. Reserves are not only money. A plan with zero schedule buffer is as fragile as one with zero cost buffer — and in a year of supply delays, schedule contingency was often the thing that actually saved delivery.
Size it from the risks, then sanity-check
The defensible way to size contingency is bottom-up from your risk register. For each significant risk, estimate the cost impact and the probability, and the expected value gives you a grounded figure to aggregate. For projects with many small risks, a quantitative pass — even a simple Monte Carlo simulation across uncertain estimates — turns a guess into a range you can defend to a sponsor. Then sanity-check the total against experience: if the model says two percent on work you know to be volatile, the model is wrong, not the world. Document the basis so that when someone asks why the reserve is what it is, you have an answer better than "it felt right."
Finally, govern the draw-down. Decide in advance who can release reserve, on what evidence, and how it is reported. With approvals now often happening across distributed teams, a clear, written rule for accessing reserve prevents both the panic raid and the reserve that nobody dares touch.
If you want a second set of eyes on how your projects estimate, hold and govern reserves, XNM's program & project delivery advisory can help you build the discipline in.