The Short Version
I've reviewed bid-to-actual reports with hundreds of builders, and the contingency story is almost always the same. The builder added 10% on top of everything. The job ran over by 12%. The contingency got burned in the first two months on two problems that, in hindsight, were completely predictable. There's a better approach. It takes the same 20 minutes as flat-percentage contingency to calculate, but it actually reflects the risk profile of the specific project type. This post walks through which cost categories carry the most variance, how to assign category-specific buffers, and how to frame contingency in your proposal so clients understand what they're getting.
Sound Familiar?
Signs your contingency approach is working against you:
- You add the same contingency percentage to kitchen remodels and new builds, even though their risk profiles are completely different
- Your contingency reserve runs out before the job is halfway done — then you're absorbing overruns out of margin
- You've been surprised by MEP rough-in costs on remodel projects more than twice in the past two years
- You've never done a formal bid-to-actual review to see which cost categories actually produce variance
- Your contingency line item disappears in the proposal — clients see the total number and think they're buying a fixed price
- When the contingency is unused, you feel like you overcharged. When it's used up early, you feel like you undercharged. Neither feeling has a data answer behind it
What We Found
Why Flat-Percentage Contingency Is a Guess, Not a Risk Management Strategy
Flat contingency — "I add 10% to everything" — is the construction industry's most common estimating habit. It also explains why so many builders are chronically surprised by overruns in the same two or three categories, job after job.
Here's the problem. The cost categories in a construction estimate don't have the same variance profile. Finish carpentry on a custom home is predictable — the materials are specified, the labor rate is known, and surprises are rare. Structural work in a remodel project is inherently unpredictable — you don't know what's behind the walls until demo day. Applying the same 10% buffer to both is mathematically equivalent to having no contingency strategy at all.
After running bid-to-actual reviews with more than 300 builders, I can tell you with confidence that 60-70% of contingency burn in residential construction concentrates in three categories:
- MEP rough-in (mechanical, electrical, plumbing) — especially on remodel projects where existing conditions are unknown until work begins
- Structural and framing — particularly on additions and renovations where existing framing, headers, and load paths aren't documented
- Site conditions and earthwork — drainage surprises, soil bearing failures, and rock that didn't show up on the geotech report
These three categories routinely run 15-25% over initial estimates. Finish work, fixtures, and cabinetry — the categories builders tend to over-contingency because the dollar amounts are large and visible — run over at less than 5% in most projects.
A data-driven contingency model uses these variances. High-risk categories get 15-20% buffers. Low-variance categories get 3-5%. The total contingency dollars land in roughly the same place as a flat 10%, but they're positioned where the money will actually be needed.
Category-Specific Contingency: How to Build the Model
Setting up a category-specific contingency model doesn't require a spreadsheet PhD. It requires three things: knowing your cost categories, having 12-24 months of bid-to-actual data, and updating the model once a year when you do your financial review.
If you don't have formal bid-to-actual data yet, start with these industry benchmarks and calibrate them to your own history over time:
High-variance categories (15-20% contingency buffer recommended):
- MEP rough-in on remodel or renovation projects
- Structural framing on additions or second-story work
- Foundation work on projects with unknown soil conditions
- Hazmat abatement (lead, asbestos) when full extent is unknown at bid time
- Earthwork and site prep on lots without recent geotech
Medium-variance categories (8-12% contingency buffer recommended):
- Exterior envelope — roofing, windows, doors, siding where existing conditions affect scope
- Interior demo on projects where the extent of damage isn't fully visible before work starts
- HVAC replacement in older homes where ductwork size and routing aren't confirmed
Low-variance categories (3-5% contingency buffer recommended):
- Finish carpentry, millwork, and cabinetry — specified and priced before contract
- Fixtures, appliances, and hard finishes — specified and priced before contract
- Painting — linear relationship between surface area and labor/material cost
- New construction framing — known scope on designed plans
The practical implementation: when you build your estimate, flag each cost line item with a contingency tier (high, medium, low). Calculate the contingency dollar amount per tier. Sum them. That's your total contingency reserve.
For a typical $400K residential remodel, this approach might produce a $38,000 contingency reserve — close to the $40,000 a flat 10% would produce, but allocated very differently. A large fraction sits against MEP and structural line items. A small fraction sits against finishes. When the MEP rough-in surprises hit (and they will), you have the reserve where it belongs.
The other thing this model does: it tells you, for any given project, whether the contingency needs are unusually high. A remodel with extensive structural unknowns and old MEP might produce a 14% total contingency rate using the category model. That's the honest number. If a client won't accept it, you need to know that before you sign the contract — not after the surprises hit.
How to Present Contingency in the Proposal Without Losing the Job
The other half of the contingency problem isn't math — it's communication. Builders who add contingency and bury it in the total number create a trust problem. Clients who see the contingency reserve itemized and unused at job close wonder if they overpaid. Clients who see it consumed in the first two months without explanation wonder what went wrong.
The approach I recommend — and the one that survives comparison shopping — is transparent contingency framing. Include the contingency as a named line item in the proposal: "Project contingency reserve: $38,000." Give it a one-sentence explanation: "This reserve covers field conditions and coordination adjustments that can't be fully scoped before work begins, specifically in MEP rough-in and structural areas. Unused contingency is returned or applied to scope additions at your direction."
This framing does three things. It educates clients about construction risk honestly, so they don't feel blindsided when the reserve gets used. It positions Go First as a transparent advisor rather than a number-padding contractor. And it protects the contingency from being negotiated away — clients who understand what it's for are much less likely to push back on removing it.
On the change order side, document every contingency draw explicitly: what condition triggered it, what work it covered, and how much remains in the reserve. A running contingency ledger in JobTread or your project management tool turns a "the money disappeared" narrative into a documented cost history. Clients who can see exactly where every contingency dollar went don't have disputes about it.
The builders I've worked with who use transparent contingency framing consistently report two benefits: fewer change order disputes (because scope additions and field conditions are handled within an expected framework), and higher close rates on competitive bids (because the honest number — even when it's higher — builds trust that less transparent proposals don't).
Most builders leave contingency as a hidden line item because they're afraid it'll hurt the bid. In my experience, the opposite is true. A builder who explains contingency clearly is the one clients trust to manage their project honestly.
Stop Guessing Where Your Profit Is Going
Download the Profit Leak Checklist used by 300+ builders. Find the exact cost codes, billing gaps, and overhead mistakes hiding in your numbers — free, instant download.
Get the Free Checklist →Frequently Asked Questions
For residential remodels and renovations, 8-15% total contingency is appropriate depending on how much of the scope involves unknown existing conditions. New construction on designed plans warrants 5-8% contingency because the scope is more predictable. The right approach is category-specific: high-variance categories like MEP rough-in and structural work on remodels should carry 15-20% buffers, while finish carpentry and specified fixtures can carry 3-5%. A flat 10% applied equally to all categories misprices risk in both directions.
Both approaches work, but a separate named contingency line item is more defensible in client communication and in change order documentation. When contingency is baked into individual lines, unused buffer disappears silently and used buffer gets confused with cost overruns. A named contingency reserve that draws down transparently gives clients a clear view of how project risk is being managed — and protects the builder from "I thought this was a fixed price" disputes at project close.
Allowances cover specified items where the exact selection hasn't been made yet — a client says "we haven't picked the tile" so you include an allowance for tile at $8/sq ft. Contingency covers unknown site or field conditions that can't be priced in advance. Allowances are client-driven selection risk. Contingency is construction-condition risk. They should be separate line items. Builders who combine them create accounting confusion when allowances run over and contingency is consumed simultaneously — it becomes impossible to distinguish selection overruns from field condition overruns.
That depends on how your contract is structured. On a fixed-price contract, unused contingency stays with the builder — it becomes additional margin. On a cost-plus or transparent open-book contract, unused contingency is typically returned to the client or applied to approved scope additions. Either approach works. What matters is that the contingency policy is documented in the contract and communicated to the client upfront. Builders who address contingency in the contract avoid disputes about it at closeout.
Annually, as part of your year-end bid-to-actual review. Pull your completed projects from the past 12 months, compare estimated contingency to actual contingency used by category, and adjust your category buffers accordingly. If your MEP rough-in contingency has run out on four of the last six remodels, the buffer is too low — raise it. If your finish work contingency is untouched on 90% of jobs, the buffer is too high — tighten it. The model should reflect your actual project experience, not industry benchmarks you've never validated against your own numbers.
