The rules of thumb
| Project type | Contingency | Why |
|---|---|---|
| Cosmetic only (paint, floors, fixtures) | 5–10% | Few unknowns — walls stay closed. |
| Kitchen or bath, walls opening | 15% | Plumbing and electrical discovery risk. |
| Full gut / layout changes | 15–20% | Every opened wall is a chance to find something. |
| Pre-1970s house | 20%+ | Galvanized pipe, cloth wiring, no insulation, surprises. |
| Anything involving 'we're not sure what's behind there' | 20%+ | You already know why. |
The percentage applies to the whole project budget — materials and labor. On a $40,000 kitchen, a 15% contingency is $6,000 held in addition to the budget, not carved out of it.
What contingency is for (and what it isn't)
- It IS for discovery: the soft subfloor under the toilet, the junction box buried behind drywall, the vent stack where the new range hood wanted to go, the "while the wall's open we really should replace that supply line."
- It IS for the market: the tile that went up 20% between quote and order, the freight surcharge, the appliance substitution when yours went on backorder.
- It is NOT for upgrades. "Since we're over budget anyway, let's do the nicer faucet" is how a 15% reserve evaporates by week two. Upgrades come out of the budget via a trade-off (nicer faucet, cheaper hardware) — the reserve stays for things you didn't choose.
Keep contingency as its own visible line, separate from every category — not padding smeared invisibly across line items. Padded budgets get spent to the padding; a visible reserve gets defended.
When the risk actually hits
Discovery risk isn't spread evenly across the project — it front-loads into demolition and rough-in, the phases where walls open and the house tells you the truth. By the time drywall closes, most of the unknown is behind you. Practical consequence: if you reach the halfway point with the reserve intact, you can consciously release part of it — that's when the nicer faucet becomes a legitimate decision instead of a leak. Track spend by phase (estimated vs. actual, per category) so "how much unknown is left" is a number you can read, not a feeling.
Old houses: believe the percentage
Pre-1970s homes earn their 20%+ the honest way: galvanized supply lines that crumble at the touch of a wrench, two-wire electrical with no ground, framing dimensions that predate standards, six layers of flooring stacked like sediment. None of these are reasons not to renovate — they're reasons the reserve is real money, decided before demo. The homeowners who get hurt aren't the ones who find problems; they're the ones who find problems with no plan for funding the fix, mid-project, with a crew on the clock. If the numbers still feel tight, the priority ladder is the tool: a smaller project fully funded beats a bigger one that stalls at 80%.