An attendance sheet answers one question precisely: who was on site, when, and for how many hours — exactly what payroll needs every month. It does not answer a different question entirely: what did those hours actually buy? A worker present for a full shift can still be only half as productive because of a material delay or a congested site, and attendance alone will never reveal that. This guide is about that second question: how productivity actually gets measured, not just attendance.
Key takeaways
- Attendance and productivity are two different questions: one prices the cost, the other prices what that cost actually bought.
- The unit worth measuring is the task, not the day — estimated hours against actual hours per line of work, with a named crew attached.
- Most paid hours do not vanish from the attendance sheet; they vanish on the way to a task — paid and logged as attendance, but never reaching a line whose productivity gets measured.
- The only comparison that means anything is a crew against its own history on similar work, not a crew against a different one doing different work.
- A bad productivity number is a question worth investigating, not a verdict — using it to punish stops honest hour-logging, which erases the very metric it depends on.
Attendance answers cost, not output
An attendance record is a precise log of a specific question: who showed up, where, and for how long. That is entirely sufficient to run payroll — but it says nothing about what got built in those hours. The gap between the two questions is the gap between 'what did we pay' and 'what did we buy with it'.
| Question | What answers it | What it doesn't answer |
|---|---|---|
| Who worked, and for how long? | An attendance record, accurate to the minute via geofencing | What that worker actually accomplished in those hours |
| What did we build against plan? | Estimated versus actual hours on a task | Exactly who, individually, was slow or fast |
A company holding only the first record knows its labour cost precisely, and has no idea whether that cost is buying shrinking output week over week. That is exactly why a project can look perfectly disciplined on its attendance sheets while quietly running over its labour budget.
The two numbers every task needs
Productivity needs two numbers at the level of the task itself, not the day and not an individual worker: the hours it was estimated to take at planning, and the actual hours the field team logged against it, with the crew assigned to it by name. The gap between those two numbers is the first real signal — not a general feeling that 'work is moving slowly'.
Choosing the task as the unit rather than the day is not a technical footnote. A worker present for a fully documented 8-hour shift may have spent half of it waiting on a material that never arrived, and the other half genuinely productive on one line item — the attendance sheet logs 8 hours of perfectly correct attendance, while the task itself logs a clear gap between estimated and actual that reveals the real problem.
A fully worked example
A formwork line item: planning estimated 8 workers × 5 days × 8 hours = 320 man-hours to complete it. The actual crew stayed 6.5 days instead of 5 to finish the same scope, with the same crew — 416 actual man-hours.
| Item | Value |
|---|---|
| Estimated hours | 320 hours (8 workers × 5 days × 8 hours) |
| Actual hours | 416 hours (8 workers × 6.5 days × 8 hours) |
| Variance | +96 hours, 30% over plan for this specific task |
| Cost impact if those hours are paid | 96 extra man-hours × the crew's hourly rate |
This variance will not show up on the monthly attendance sheet — the crew was fully present, the work got finished, and payroll runs without a single flag. It only shows up when estimated hours are compared against actual at the level of the task itself, which is exactly what makes this comparison a separate tool from attendance, not a replacement for it or a duplicate of it.
Why the hours go missing before they ever reach a task
The most common failure is not that hours go unrecorded — it is that they are logged as attendance and never reach a task whose productivity gets measured against it. Attendance works, payroll runs correctly, and productivity stays completely invisible even though cost is fully visible.
- General or support labour never assigned to a specific task, so its cost sits outside any productivity measure entirely.
- Workers moved between tasks more than once a day with no record of actual time on each move.
- A task closed administratively while some crew members are still finishing its remaining details, with no update to the final hours.
- A subcontractor managed on a lump sum per line with no hour breakdown at all, so there is no data to compare in the first place.
Benchmark a crew against itself, not against another
Comparing a finishing crew to an excavation crew means nothing — the nature of the work is completely different. The only comparison that means anything is the same crew, or the same trade, on the same type of work across different periods or different sites.
Log consistently
Document estimated and actual hours for every task of its kind regularly, not only once you suspect a problem.
Accumulate at least two cycles
One variance could be a passing circumstance; a pattern repeating across two or three cycles is what actually deserves a decision built on it.
Compare like for like
Compare the same crew on this month's formwork line against its own performance on the same line last month, not against a different crew building a different scope on a different site.
Investigate before you conclude
A repeated variance could mean a genuinely slow crew, or a task whose hours were misestimated from the start at planning — and the fix is completely different in each case.
What a productivity number is actually good for
A bad productivity number is the start of a "why" question, not a ready verdict on a crew or an individual. Rework from poor quality, weather that wiped out half a shift, a crew sized wrong for the scope, or a delayed material that kept the crew present and logged with no real work to do — all of these produce the exact same variance on screen and each needs a completely different response.
Using the number for direct individual accountability before investigating it first produces a well-known backfire: the crew stops logging hours honestly, and the very signal that was supposed to help everyone disappears. The goal is improving planning and execution together, not assigning blame to whoever logged the number honestly.
How this works in muqawil
Every task in muqawil carries an estimated hour figure set at planning and an actual hour figure the field team logs against it, with the crew assigned to it by name rather than a bare headcount. That puts the comparison between planned and actual at the level where it means something — the task itself — not only at the level payroll needs, which is the day.
That data is completed by what sits next to it: daily geofenced attendance gives the cost side and the paid hours precisely, and the bill-of-quantities line a task is building toward gives the output side — so the two questions that usually stay separate finally meet: what did we pay, and what did we actually buy with it.
Frequently asked questions
What is the difference between attendance and labor productivity?
Attendance logs who worked, when, and for how many hours — which prices the cost. Productivity compares those hours against what was estimated to finish the same scope of work — which tells you whether that cost is buying shrinking output or not. The first is essential for payroll; the second is essential for knowing where the time actually goes.
What is a normal productivity variance to expect?
There is no single figure that fits every type of work; it depends on the nature and complexity of the task. What matters is not comparing the variance to an absolute number, but comparing it to the same crew's own performance on the same work in earlier cycles — a repeating pattern is the signal, not one isolated figure.
Should productivity be tracked per worker or per task?
At the level of the task and its assigned crew, not usually an individual in isolation. Most construction work is collective by nature, and separating one worker's contribution from another's inside the same crew is harder to build a reliable measure on, while a task has clear boundaries with a defined plan and actual.
How do weather or a material delay affect a productivity number?
They produce a clear gap between planned and actual even though the crew may have been fully present and done nothing wrong. That is exactly why the number is a starting point for asking why, not an automatic verdict on crew performance — a cause outside their control needs documentation, not punishment.
Is it appropriate to use productivity tracking to discipline workers?
It is not recommended. Using the number for direct individual accountability pushes a crew to stop logging hours honestly, which erases the metric itself. The correct use is improving future hour estimates and understanding the causes of variance, not assigning blame to whoever logged it honestly.
Compare planned and actual hours at the level of every task
Estimated and actual hours per task, with a named crew attached, alongside daily geofenced attendance — so you know not just what you paid, but what you actually bought with it.