The construction labor rates for estimating that hold up are hours per unit taken from your own finished jobs — something like 0.12 hours per linear foot of interior partition framing — not a number from a cost book or the one a foreman remembers. A book rate describes an average crew under average conditions that somebody else measured. Your crew, your sites and your definition of done are what you are actually pricing. Getting those numbers takes two things most builders already half-have: hours coded to cost codes and installed quantities recorded against the same codes.
A labor rate is two numbers, and only one of them moves
The wage rate is dollars per hour. It is base pay plus burden — payroll taxes, workers' comp, benefits and any paid time that isn't on a task. It comes from payroll, changes once or twice a year and rarely sinks a job on its own.
The production rate is hours per unit. Hours per square foot of drywall hung, per linear foot of base installed, per door hung and trimmed. This is the number that swings with crew, site and scope, and it is where estimates go wrong. Labor cost on a line item is quantity times hours per unit times burdened wage. Keep the two inputs separate in the estimate so each can be checked against its own source.
Keep production rates in hours, not dollars. A dollars-per-foot rate from two years ago has old wages baked into it. If the next job runs over, there is no way to tell whether the crew was slower or the raise you gave last spring is showing up. Hours survive raises. Company overhead does not belong in the wage either — it is a separate number, and your overhead rate belongs to your company, not to each labor line.
Why book rates and remembered rates describe someone else's crew
- Cost book rates — measured on other crews and averaged across regions and job types. The unit definition may or may not include layout, material handling, cleanup or setup. The book knows nothing about your access, your travel time or what it means to frame inside an occupied house.
- Remembered rates — built from the best day on the best job. Memory keeps the afternoon a wall went up fast and drops the morning lost to a late delivery, the return trip for punch and the second mobilization. Remembered rates run fast, which is why they win bids and lose money.
- A sub's quote or a competitor's number — tells you what the market will bear for the work. It says nothing about what your own crew will take to do it.
None of this makes a cost book useless. For work you have never self-performed, it is a starting point. Use it, mark the line as a book rate and capture your own hours the first time you do the work, so the book rate gets replaced instead of copied forward for five years.
Building hours per unit from job costing
The arithmetic is division: actual hours charged to a cost code divided by the quantity installed under that code. Everything else is about making sure both numbers are true. If you are not yet tracking hours by code, start with how construction job costing works — this method sits on top of it.
A worked example, with illustrative numbers. On one remodel, the interior wall framing code shows 212 hours against 1,840 linear feet installed — 0.115 hours per foot. A smaller job shows 96 hours on 760 feet — 0.126. A second-floor addition where every stick was carried up a stair shows 140 hours on 980 feet — 0.143. Three jobs, three numbers and one obvious reason the third is higher.
- Close out hours by code after the last time entry is approved, not while the job is running and entries are still coming in.
- Record installed quantity per code — the takeoff quantity adjusted for change orders, or a field count. It has to be what was built, not what was bid.
- Divide hours by quantity and write the conditions next to the result: occupied home, second floor, winter, new lead carpenter, tight access.
- Line up the same code across three to five finished jobs.
- Pick the rate for the next estimate and note which jobs it came from and why.
The rates only compare if the codes mean the same thing every time
The division is easy. The hard part is that the hours and the quantity have to mean the same thing on every job. If framing includes blocking and backing on one job and blocking is its own code on the next, the two rates are measuring different work. The fix is a cost code list written down once and used on every job — the post on setting up cost codes for construction covers how to structure one. For production rates specifically, these rules matter most:
- One scope per code, written down — what is in and what is out, so a new foreman codes the same way an old one does.
- One unit per code, fixed — linear feet, square feet or each. Switching units between jobs breaks every comparison.
- Time entered against a code daily. Hours reconstructed on Friday get spread across whatever codes seem plausible.
- Change order labor coded to the change order. Folded back into the base code, it makes the base rate look slower than it was.
- No catch-all labor or general code that quietly absorbs hours nobody wanted to assign.
Even with clean codes, three jobs will give three numbers. Don't average them blindly — the spread is information. When the numbers sit close together, use the middle one. When one is an outlier with a known cause, like the carried-upstairs addition, leave it out of the base rate and price that condition as an adjustment the next time it shows up. When the numbers are far apart and nobody can explain why, suspect the coding before the crew. Check whether the code covered the same scope on each job.
Conditions that reliably slow work — occupied homes, height, poor access, cold weather — should be priced as named adjustments to the line, not by inflating the base rate for every job. Things that might happen but aren't known yet are a different matter. Those belong in a contingency that names the risks it covers, where they can be defended when the owner asks what the number is for.
Closing the loop while the job is still running
Rates only get better if every finished job feeds back into the next estimate. The earlier signal comes during the job: budget vs. actual by cost code shows a line drifting while there is still time to ask why. When the budget was built from the estimate, the comparison is line to line, and a framing overrun shows up as a framing overrun rather than a vague sense that labor is running heavy.
In BuilderSolve, the estimate is built line by line on a shared rate library, and the budget and schedule come from that estimate. Time entries land against those lines, and budget vs. actual shows each one. The AI assistant can draft time entries for a person to approve or reject — it does not approve them. What the platform does not do is decide your new production rate for you. An outlier that should be excluded looks identical to a formula as one that should count, so updating a rate in the library is a call a person makes after reading the jobs. For builders without an estimator on staff, the BuilderSolve estimating team can run that close-out review and keep the library current.
What to do this month
- Pick the five line items where labor is the largest share of your estimates. Framing, drywall, trim and demo are common, but use your own.
- Pull the last three to five finished jobs that included them. If the hours were never coded, accept that you don't have a rate yet — don't invent one from the payroll total.
- Compute hours per unit for each job and write down the conditions beside each number.
- Set the rate in your estimating library with a note of which jobs it came from.
- On the next job, enforce written scope definitions and daily coding for those five codes only. Add more codes once those five are clean.
After two or three jobs coded this way, the rate on those lines stops being a guess borrowed from a book. It is a measurement of your crew, and it gets more accurate with every job you close out.
Common questions
What are typical construction labor rates for estimating?
There is no typical rate worth estimating with — the useful number is hours per unit from your own finished jobs. Wage rates come from your payroll plus burden. Production rates from cost books are averages across other crews and conditions. Use a book rate only for work you have never done, mark it as such and replace it with your own hours after the first job.
How do I calculate labor production rates from past jobs?
Divide the actual hours charged to a cost code by the quantity installed under that code. For example, 212 hours on 1,840 linear feet of framing is about 0.115 hours per foot. Do it after the job closes, use installed quantity rather than bid quantity and note conditions like access or occupancy beside the result.
Should labor rates in an estimate include overhead?
No — the labor rate should include wage and burden but not company overhead. Burden covers payroll taxes, workers' comp and benefits tied to the hours worked. Office rent, admin salaries and vehicles are company overhead and should be recovered through a separate overhead rate, so labor lines stay comparable across jobs and overhead stays visible.
How many jobs do I need before my own labor rates are reliable?
Three to five finished jobs per cost code is enough to see a pattern and spot outliers. One job gives a number but no way to tell whether it was normal. What matters more than the count is consistent coding — five jobs coded differently are worth less than three coded to the same written scope and unit.
What if I have no job cost history to build labor rates from?
Start with a cost book rate for each line, flag it as unverified and begin coding hours daily on your next job. Pick the five line items with the most labor first rather than coding everything at once. After two or three jobs those lines carry your own rates, and the book numbers can be retired.