On this page
- Step 1: Pick the cost codes that matter
- Step 2: Set the budget quantities and unit rates from the estimate
- Step 3: Code hours to the same cost codes
- Step 4: Record quantities installed, with rules of credit
- Step 5: Calculate earned hours and the productivity index
- Step 6: Forecast hours at completion
- Step 7: Talk to the foreman, then feed the forecast
- Worked example: branch conduit over six weeks
- Common mistakes
- How os.construction handles it
- FAQ
To track labor productivity, set a budget unit rate for each major labor cost code (budget hours ÷ budget quantity), record hours and quantities installed against the same codes each week, and compare them: earned hours are the quantity installed times the budget unit rate, and earned hours ÷ actual hours is the productivity index. Then forecast the hours to finish from the remaining quantity and a unit rate you can defend.
A labor cost report that only compares hours spent with hours budgeted can’t tell a crew that’s ahead of the work from one that’s behind it. Fifty percent of the hours gone sounds fine until you learn forty percent of the work is done. Quantities installed are the missing half, and once you have them, the forecast stops being a guess. It’s also the field version of the earned value idea behind a performance factor: measure what was earned, not just what was spent.
Step 1: Pick the cost codes that matter
Don’t track every code. Choose the ones that carry most of the labor budget and have a clear unit of measure: linear feet of conduit or pipe, square feet of board or deck, cubic yards placed, fixtures set, devices installed. On most self-perform jobs a handful of codes account for most of the labor risk.
Step 2: Set the budget quantities and unit rates from the estimate
For each tracked code, carry the estimate’s quantity and hours into the budget:
budget unit rate = budget hours ÷ budget quantity
If the estimate was built from production rates, this is that rate. If it wasn’t, now is the time to work it out, before the first timecard. When the quantity changes through a change order, update both the budget quantity and the budget hours, or every number below drifts. How to build a construction estimate covers where those rates come from.
Step 3: Code hours to the same cost codes
Foremen code their crews’ hours daily to the tracked codes, at the same level of detail as the budget. Hours dumped into a general code can’t be compared with anything. If a crew splits its day between two codes, split the hours too; an honest estimate of the split is better than all of them on one line.
Step 4: Record quantities installed, with rules of credit
Each week, record the quantity in place for each tracked code, cut off on the same day as the hours. Quantities in the daily report make the weekly count easier and give you something to check it against.
For work with distinct steps, agree rules of credit up front, so partial work earns partial quantity. For example, a drywall team might credit a square foot as 50% when hung, 80% when taped and 100% when finished. The split is yours to set; what matters is that it’s written down and used the same way every week.
Step 5: Calculate earned hours and the productivity index
earned hours = quantity installed to date × budget unit rate productivity index (PI) = earned hours ÷ actual hours actual unit rate = actual hours ÷ quantity installed
A PI of 1.0 means the crew is on the estimate. Below 1.0 it’s using more hours than the estimate allowed for the work done. Some companies invert the ratio; pick one convention and label it on every report. If you’re used to cost-based earned value, PI is the hours version of CPI.
Look at the week as well as the job to date. The cumulative PI moves slowly; the weekly PI shows what’s happening now.
Step 6: Forecast hours at completion
hours at completion = actual hours to date + remaining quantity × forecast unit rate
The forecast unit rate is a decision, not a formula. Use the budget rate if the cause of the early overrun is behind you, the to-date rate if nothing has changed, or the recent rate if the crew has settled into a steady pace. Write down which one you used and why. Multiply the hours by the burdened labor rate to turn them into cost; the labor burden calculator gives you that rate.
Step 7: Talk to the foreman, then feed the forecast
A low PI is a question, not a verdict. Ask the foreman what’s slowing the crew: access, missing material, rework, design questions, trade stacking, crew size, or a quantity that was simply wrong in the estimate. Some causes can be fixed this week, and some belong in a change event because someone else caused them. Then carry the hours forecast into the cost-to-complete review, so the margin forecast reflects it this month.
Worked example: branch conduit over six weeks
An electrical contractor budgets 12,000 linear feet of branch conduit at 1,800 hours, a budget unit rate of 1,800 ÷ 12,000 = 0.15 hours per foot. The burdened labor rate is $62 an hour. After six weeks:
| Week | Feet installed | Hours | Unit rate (h/ft) | Earned hours | Weekly PI |
|---|---|---|---|---|---|
| 1 | 500 | 120 | 0.240 | 75.0 | 0.63 |
| 2 | 800 | 160 | 0.200 | 120.0 | 0.75 |
| 3 | 950 | 170 | 0.179 | 142.5 | 0.84 |
| 4 | 1,050 | 170 | 0.162 | 157.5 | 0.93 |
| 5 | 900 | 150 | 0.167 | 135.0 | 0.90 |
| 6 | 1,000 | 150 | 0.150 | 150.0 | 1.00 |
| To date | 5,200 | 920 | 0.177 | 780.0 | 0.85 |
To date the crew has earned 5,200 × 0.15 = 780 hours for 920 spent, a PI of 780 ÷ 920 = 0.85. By hours, the code looks 920 ÷ 1,800 = 51.1% complete; by quantity it’s 5,200 ÷ 12,000 = 43.3% complete. A report that used hours as progress would overstate this code by almost eight points.
The foreman explains the first two weeks: the first floor had the most complicated layout and the crew was waiting on a coordination answer for the corridor ceiling. Since week 4 the crew has run close to the estimate. Three forecasts for the remaining 12,000 − 5,200 = 6,800 feet:
| Forecast basis | Rate (h/ft) | Remaining hours | Hours at completion | Over budget | At $62/h |
|---|---|---|---|---|---|
| To-date rate, 920 ÷ 5,200 | 0.177 | 1,203 | 2,123 | 323 | $20,026 |
| Last two weeks, 300 ÷ 1,900 | 0.158 | 1,074 | 1,994 | 194 | $12,028 |
| Budget rate | 0.150 | 1,020 | 1,940 | 140 | $8,680 |
The remaining hours use the unrounded rates: 6,800 × (920 ÷ 5,200) = 1,203.1, and 6,800 × (300 ÷ 1,900) = 1,073.7. The dollar column is the rounded hours over budget times $62.
Even the budget-rate forecast is 140 hours over, because the early weeks can’t be earned back. The PM uses the two-week rate, 1,994 hours, $12,028 over the labor budget, and notes why: the cause of the slow start has passed and the crew has run within about 10% of the estimate for three weeks. Part of the early overrun traces to the late coordination answer, so the PM also logs it as a change event to see whether any of it is recoverable.
Common mistakes
- Using hours spent as percent complete. It hides an overrun until the end.
- Quantities and hours cut off on different days, which makes the weekly numbers swing for no reason.
- Hours coded to a catch-all code, so the tracked codes are missing hours.
- No rules of credit, so partially finished work is counted differently every week.
- Not updating budget quantity and hours for change orders, which makes the PI drift.
- Forecasting with the budget rate by habit, when the crew has never achieved it.
- Treating the number as a verdict on the crew instead of a question about the conditions.
- Tracking everything, so nobody looks at the few codes that matter.
How os.construction handles it
The Daily Log records manpower as workers and hours by company, cost code and location, and has a Quantities section for work in place with its unit, cost code and location, so hours and quantities are captured on the same codes in the field. Approved timesheets flow into Accounting Payroll, which posts labor and burden to the job by cost code. In the Schedule, an activity’s percent complete can be measured in units, and stored period performance records each period’s actual units, earned value and planned value for every activity. For the field side of the product, see construction daily log software.
FAQ
How do you measure labor productivity in construction?
Divide the labor hours spent on a cost code by the quantity installed, such as hours per linear foot or per square foot, and compare that unit rate with the rate the estimate assumed. Or work out earned hours, the quantity installed times the budget unit rate, and divide them by the actual hours. Above 1.0 the crew is beating the estimate; below 1.0 it's behind.
What are earned hours?
Earned hours are the hours the estimate allowed for the work actually installed: quantity installed times the budget hours per unit. If the estimate allowed 0.15 hours per foot and the crew has installed 5,200 feet, it has earned 780 hours, whatever it actually spent.
How often should productivity be tracked?
Weekly is common for the cost codes that carry most of the labor budget, with quantities and hours cut off on the same day. Daily quantities from the field make the weekly number easier to trust. Monthly is usually too late to change anything on a short activity.
Why not just use percent of hours spent as percent complete?
Because hours spent measure input, not output. If a crew has used 51% of its hours but installed 43% of the quantity, cost-based percent complete says the work is further along than it is, and the overrun only shows up near the end.
How do you forecast labor hours to complete?
Multiply the remaining quantity by the unit rate you expect for the rest of the work. That might be the budget rate if the cause of the overrun is fixed, the to-date rate if nothing has changed, or the recent rate if the crew has settled in. Add the actual hours to date to get hours at completion, and write down which rate you used and why.
Share this article