CalcStack — free, instant, no signup. The interactive calculator loads automatically.
The same $480k project carries $120,000 of contingency at schematic and $38,400 at construction documents. Contingency shrinks as information grows — no line at all claims certainty you do not have.
Contingency is the most misunderstood line in estimating — treated as padding to hide or profit to harvest, when it is actually uncertainty priced. The discipline: contingency tracks information. At schematic, a $480,000 estimate carries 25% ($120,000); at design development, 15% ($72,000); at complete construction documents on a known site, 8% ($38,400) — with tiers added for unknown site conditions and volatile material markets. An estimate with no contingency line is not aggressive, it is claiming certainty the documents do not support. This calculator sizes the line by stage and risk — and the FAQ covers how to draw it down honestly.
Estimators use it before the review: the same $480k project carries $120k of contingency at schematic and $38k at construction documents — contingency is uncertainty priced, and it shrinks as information grows.
Three different lines covering three different unknowns, and conflating them is how estimates go soft. Contingency covers UNCERTAINTY in what is being built — design incompleteness, quantity risk, coordination gaps between trades; it shrinks as documents improve, which is the entire logic of the staged percentages. Allowances cover KNOWN items with unknown selections — the owner has not picked the tile, so the estimate carries $8/sqft for tile; allowances are placeholders for decisions, listed item by item with the unit basis written down, and they reconcile to actual selections by change order. Escalation covers TIME — material and labor price movement between estimate day and purchase day; in stable markets it is a small percentage, in volatile ones (lumber 2021, copper, switchgear lead times) it deserves its own clause: price-at-time-of-purchase with index documentation is fairer to both sides than burying a guess in contingency. The owner-side mirror: owners carry their OWN contingency (5–10% new build, 10–20% renovation) for scope changes and soft costs — separate from the contractor's — and projects where only one side carries contingency end in the funding crisis the other side predicted. Label all three lines separately in the estimate; the transparency is what makes the number defensible at review.
With the drawdown schedule and the allowance schedule — transparency is the defense. The argument that works: contingency is not a margin line, it is an uncertainty line, and it comes with a written management plan — it is drawn against identified risks (list them: design completeness at the current stage, site unknowns, long-lead price risk), it is tracked separately in job cost, and unspent contingency returns to the owner or rolls forward by agreement. The reframe that lands: cutting contingency does not remove the risk, it removes the budget for the risk — the unknowns remain, and they will surface as change orders at worse pricing with worse timing and worse feelings. Owners have lived this: the project that cut contingency to hit a number and then exceeded it in change orders is a universal story. The collaborative structure: shared savings clauses (unspent contingency splits 50/50) align incentives — the contractor is not tempted to spend it, the owner recovers the residue. What to watch from the other side: contingency as hidden profit — a contractor whose contingency is never reported, never reconciled, and always fully consumed is billing the line, not managing it; monthly contingency-versus-consumed reporting (the owner can ask for this) keeps the line honest. And calibrate from history: your own closed jobs — contingency carried versus consumed, by project type — are a better table than any industry percentage.
Renovation is the site-unknowns tier in this calculator for a reason: the building is concealing its conditions until demolition, and no document set fixes that. The specific unknowns: concealed utilities that are not where the as-builts swore (as-builts lie — they document intent, not reality), structural surprises (the beam that is not sized as drawn, the slab that is thinner), hazardous materials (asbestos and lead in pre-1980 buildings — testing BEFORE the estimate converts the biggest unknown into a priced line), code-triggered scope (opening walls triggers insulation, accessibility, and fire-alarm upgrades the program never wanted), and occupied-building logistics (night work, phasing, protection — productivity penalties that belong in the base, not contingency). The practices that shrink renovation contingency honestly: selective demolition and investigation BEFORE final pricing (open the ceiling, scan the slab, test the materials — $5–15k of pre-construction investigation routinely removes $50k+ of contingency), unit prices pre-agreed for the predictable unknowns (per yard of unsuitable soil, per square foot of unforeseen repair), and a clear allowance/contingency split so the owner sees which unknowns are priced and which are budgeted. The honest number: renovation contingency at 10–20% even with good documents is not conservatism — it is actuarial reality, and the contractors who carry it finish with the owner relationship intact.