You avoid cost overruns in construction by pricing the whole job before you sign anything, then watching the numbers that move before the money does: what you sign against what you priced, the price per unit on the first deliveries, how fast the contingency goes against the work, and the finish date. Actual spend passing the estimate is the last signal, not the first. By then the money is spent or signed for, and the only choice left is where it comes from.
On one line, an overrun is the work costing more than you estimated for that same work. The one you're trying to avoid is the job's: overruns adding up past your contingency and into your profit. Spend measured against a whole scope's estimate tells you nothing until you know how much of the work that spend bought.
Why jobs run over
- The estimate missed something: a line nobody priced, a quantity measured wrong, a bid against a scope nobody wrote down.
- The house was hiding something: rot under the tub, wiring that won't pass, a joist somebody notched for a drain.
- Work got added and nobody wrote a change order for it.
- The price per unit moved: the supplier went up between your quote and your order, or a lumber quote expired.
- You picked nicer than you priced: the tile, the faucet, the cabinet line. That's a choice, so price it as one.
- Material went in the dumpster: bad cuts, breakage, a second order of the same thing.
- The job ran long. If you own the house, interest (if you borrowed), taxes, insurance, utilities and upkeep run every day it's yours, and no scope's budget shows them.
A lot of it is decided before you sign
- Write the scope of work before you ask for a price, and give every bidder the same one. A bid against a vague scope leaves gaps, and every gap is a change order waiting to be written.
- Ask for unit prices on what you can't see yet — per sheet of subfloor, per foot of rotten sill — so the surprise behind the wall has a price before anyone finds it.
- Compare each bid with what you priced that work at before you sign. The first signal below is that comparison.
- Carry a contingency: money set aside on top of the estimate for what you can't pin down yet — what's behind the walls, prices that move, the line you missed — that you expect, taken together, to cost you something. It's meant to be spent. How much to hold is its own question. Here, it's a number to watch.
The early signals, in the order they show up
Each of these can tell you a scope is going over long before its spend passes its estimate. They're in the order they tend to appear on a job.
1. You signed for more than you priced
A signed price is money you've agreed to spend; the bill just hasn't come. Say you priced the kitchen cabinets at $9,000 and the shop's price after field measure is $10,400. That's $1,400 over, decided at the signing table, long before the last invoice.
Compare each contract with what you priced that work at. If you priced the kitchen by line, compare by line, because the scope's total hides an overrun behind the lines nobody has signed yet. Say the kitchen is estimated at $24,000 and you've signed four contracts for $19,400. By the scope's total you're $4,600 under. But those four lines were priced at $17,300, so you've signed $2,100 more than you priced them at. What you've signed for, paid or not, is what you've committed: the middle of the estimated, committed and actual columns in a flip budget.
2. The price per unit drifts on the first deliveries
For what you buy yourself, or what a sub prices by the unit, divide what each delivery cost by the quantity it covered. Say you priced 200 square feet of tile at $11 a square foot, $2,200: 100 for the floor and 100 for a full-height backsplash. The floor tile comes to $1,450: $14.50 a foot. The line still looks fine, $1,450 spent of $2,200. But that 100 feet was priced at $1,100, so it's already $350 over on the tile you've bought, and if the wall tile comes in at the same price the line ends at $2,900, $700 over. You know that while the wall tile isn't bought yet, and a cheaper one there is still an option.
A lump-sum contract doesn't drift this way: its price moves by change order, or by an escalation clause if you signed one, not by delivery. Your own crew does, with the hour as the unit. Say a bathroom floor on another job was priced at 30 hours of tile setting, and the crew is 20 hours in with a third of it down: they're on pace for 60.
3. You're buying past the takeoff
The takeoff is the quantity you measured, from plans or the house itself, to price the job. Buying a little past it for cuts and breakage is normal, and you price that in. A third trip for floor tile you measured at 100 feet is something else: the price didn't change, the cuts, the breakage or the measurement did, or the tile went to another job. And the third batch may not match the first two. That's a different conversation from a price problem, with a different person.
4. Contingency goes faster than the work
Compare the share of contingency you've used with the share of the work that's done. Say the kitchen is one scope of a rehab with $75,000 of scoped work and $6,000 of contingency, and the job is 40% done. If overruns and surprises across the job have used $3,300 of the contingency, that's 55% of it gone at 40% of the way. At that pace it runs out about 73% of the way through, and everything after that comes off your profit.
Expect it to run a little ahead early, because demo is when a house shows you what it was hiding. A workable test: once the job is at least a tenth done, look hard when the share used runs more than 10 points ahead of the share done. $3,300 is 15 points ahead. $2,700, 45%, is 5 points ahead: worth watching, not yet a reason to cut.
Measure the work done by what's in place: rooms finished, tasks signed off. If you use money instead, cap each scope's spend at its estimate, or an overrun counts as progress, and know that a deposit counts as progress before anything is installed.
5. The finish date slips
Every day you own the house runs up holding costs — interest if you borrowed, taxes, insurance, utilities, upkeep — and a slip on the critical path, the chain of tasks that sets the finish date, moves your finish and the day you can sell with it. Say each extra day of owning costs you $100 and the electrician pushes a week on the critical path: 7 × $100 = $700, and it shows up on no scope. A slip on a task with float doesn't move your finish, so it adds no holding cost, until the float runs out. It can still push the next trade.
6. Actual passes the estimate
The one everyone watches, and the last to show up. In the kitchen below, spend doesn't pass the $24,000 estimate until there's only $2,800 left to pay, near the end of the job. The whole $2,800 overrun was knowable long before that.
A scope over its estimate isn't a job over budget
Three numbers get mixed up here:
- The estimate is what you priced the work at, scope by scope.
- Contingency is what you set aside on top of it for what you can't pin down yet.
- The construction budget is the two together: the job's whole number.
A scope going over its estimate isn't a crisis by itself; that's what the contingency is for. The job is over its construction budget when the overruns, added up, have used the contingency too.
Count overruns the day you see them, and underruns the day the work is done. If the roof is coming in $2,000 under, that money isn't free until the roof is finished. Spend it on the kitchen early and you've spent money the roof may still need.
To see where a scope is heading, you need two more numbers:
- Cost to complete, also called estimate to complete: everything you still expect to pay to finish the scope, from bills for work that's already done to the work that isn't. On a lump-sum contract, that's what's left to pay on it, plus any change order you expect to approve. On a unit-price one, it's what you owe for what's done plus the quantity left times the price. Everywhere else it's a fresh look at what's left, not the estimate minus what you've spent.
- Estimate at completion: what you've spent plus the cost to complete. That's your forecast of what the scope will have cost when it's done. Compare it with the estimate and you have the overrun you're heading for, while there's still work left to change.
One scope, worked
Say the kitchen on a flip is estimated at $24,000, and partway through four lines are under contract: the cabinets signed at $10,400, the countertops at $3,900, the electrical at $2,800, and the plumbing at $1,900 plus a $400 change order for a rotten drain line. The floor tile is bought at $14.50 a foot; the wall tile isn't yet. Here's where it stands:
| Line | Estimate | Spent | To complete | At completion |
|---|---|---|---|---|
| Cabinets | $9,000 | $5,200 | $5,200 | $10,400 |
| Counters | $3,600 | $0 | $3,900 | $3,900 |
| Electrical | $2,800 | $1,400 | $1,400 | $2,800 |
| Plumbing | $1,900 | $1,150 | $1,150 | $2,300 |
| Tile | $2,200 | $1,450 | $1,450 | $2,900 |
| Tile setting, your crew | $1,800 | $0 | $1,800 | $1,800 |
| Drywall and paint | $2,700 | $0 | $2,700 | $2,700 |
| Kitchen | $24,000 | $9,200 | $17,600 | $26,800 |
Read the spend column alone and the kitchen is fine: $9,200 of $24,000 spent, $14,800 left. But the cost to complete is $17,600, not $14,800, and the kitchen is heading for $26,800, $2,800 over. Every dollar of that was knowable early:
- $1,700 the day the cabinet and countertop contracts were signed;
- $400 the day you approved the drain line;
- $700 on the first tile delivery.
That $2,800 comes out of the contingency if there's room, which is the moment to check the share used against the share done, across the whole job.
Firmabuilt watches several of these for you, and two of them matter in this kitchen. The day a scope's signed contracts, change orders included, add up to more than its estimate, it raises a critical alert — "Kitchen is over before work starts" — on the job, on the alerts page and in the daily email. It reads the scope's total, so the cabinets here, over inside a kitchen that's still under, are yours to catch. And once you set the job's construction budget, contingency included, a bar on the job puts the share of contingency used beside the share of scoped work spent, and warns on the test above.
What to do when one fires
Catching it early is only worth something because you still have choices:
- Before you sign: rebid, re-spec (a different cabinet line, stock sizes), or take the price and cut the same amount somewhere you haven't signed yet.
- When a unit price drifts: pick a different material for the rest, shop the supplier, or accept it and move the money on purpose.
- When you're buying past the takeoff: walk the job with the installer and check the measurement before the next order.
- When someone wants to add work: write the change order before the work, with its price and its days.
- When contingency runs ahead: decide now what you'll give up — the appliance upgrade, the nicer wall tile — while it's still yours to decide.
- When the finish slips: pay to go faster only on the critical path, and only when the days it takes off the finish × your daily holding cost beat what it costs. Squeeze one task too far and another chain sets the finish.
Each of those choices is gone once the work it covers is in.
Questions
What is committed cost in construction?
Money you've agreed to pay under signed contracts, purchase orders and approved change orders. The term gets used two ways, though. Some reports count the whole signed amount, paid or not; others count only the part you haven't paid yet, so it shrinks every time you pay an invoice. To catch an overrun, compare the whole signed amount with what you estimated for the same work, line by line. If a report shows only the unpaid part, add back what you've already paid on those contracts first.
What's the difference between estimate to complete and estimate at completion?
Estimate to complete is what's left to pay: bills for work already done, plus what you now expect the rest of the work to cost. Estimate at completion is the whole thing: what you've spent so far plus the estimate to complete. On a kitchen estimated at $24,000 with $9,200 spent and $17,600 to complete, that's $26,800 at completion. Compare that last number with the estimate, and the gap is the overrun you're heading for.
Is a change order an overrun?
It depends on why it was written. If you're the one paying, a change order for something nobody could see when the job was priced, like a rotten drain line behind the cabinets, is what contingency is for: it puts plumbing over its estimate, and the contingency covers it. A change order for work you chose to add is a change to the plan: raise the estimate with it, so the scope still tells you how the original work is going. Either way it raises what you've committed the day you approve it. More on what a change order is and how to keep a log of them.