How Much Contingency Does Your Project Actually Need?

6 min read

Contingency gets treated as either a slush fund or an embarrassment. It is neither. It is a priced estimate of what you do not yet know, and it should shrink on a schedule as unknowns resolve.

Three separate contingencies, three separate owners

Mixing these into one number is how owners lose track of their budget.

Design contingency covers documents that are not finished. It belongs to the design phase and should decline as drawings develop — highest at concept, near zero at permit set.
Construction contingency covers execution unknowns: differing site conditions, coordination gaps, weather. It is often held by the contractor within a GMP and governed by the contract.
Owner contingency covers owner-driven change: scope additions, decisions you have not made, timing risk. It is yours and should not be visible to the contractor as available money.

Sizing by risk, not by habit

A flat percentage applied to every project is a guess wearing a uniform. Size contingency against the specific risks in front of you.

Drivers that push contingency up:

Renovation or adaptive reuse, especially with unknown existing conditions.
Incomplete or fast-tracked documents.
Long lead items in a volatile market.
Complex site conditions: poor soils, high water table, fill requirements, constrained access.
Extended schedules with escalation exposure.
Unfamiliar jurisdiction or an unusual approval path.

Drivers that allow contingency down:

Complete, coordinated documents.
Repeat prototype the team has built before.
Locked scope and an experienced, engaged owner decision maker.
Bought-out long lead packages with fixed pricing.

Set your number by listing the top risks, estimating a range for each, and summing — then sanity check against a percentage. If your bottom-up number and your percentage disagree badly, one of them is wrong and it is worth finding out which.

Escalation is not contingency

Escalation is a forecast of price movement over your procurement window. It is a line item with a defensible basis, not a risk allowance. Carrying escalation inside contingency hides both numbers and guarantees an argument later.

Rules for spending it

Every draw requires a written justification tied to a specific cause.
Owner-elected scope additions never come from construction contingency.
Track drawdown against percent complete. Spending sixty percent of contingency at thirty percent complete is a project-level warning, not a line-item issue.
Publish the remaining balance in every monthly report. Contingency that is not reported is contingency that is already gone.
Release contingency deliberately at milestones — after buyout, after foundations, after dry-in — rather than letting it drift into the general budget.

The drawdown curve

Plot expected contingency remaining against schedule and compare actuals monthly. The shape tells you more than the balance: a steep early decline signals document or condition problems; a flat curve followed by a cliff signals unrecognized issues being deferred.

The takeaway

Contingency should be built from named risks, split into three separately governed buckets, reported every month, and released on milestones. Owners who manage it this way find problems while they are still small — and finish projects with money left instead of explanations.