Problem: quotes at 30% margin, year-end accounts at 18%. Where did the money go?
Solution: hours tracked per job + weekly comparison of actual vs budgeted hours.
Result: 16.100 € in margin recovered per year, just by eliminating dead time.
The margin that evaporates without you noticing
A window installer in Northern Italy with two 3-person teams called me because the numbers weren't adding up. He quoted jobs at 30% margin. At the end of the year his accountant put a different number in front of him: actual margin 18%.
Twelve points of difference. On a revenue of 750.000 € that's around 90.000 € evaporated. Not stolen, not wasted on reckless spending. Just disappeared across a thousand small streams nobody was measuring.
First question: "Do you track actual hours per job?" Answer: "Yes, the site manager fills in the report." Second question: "Do you compare them with the budgeted hours?" Silence.
Why the report isn't enough
The site manager fills in the report at 6:30pm, after 9 hours of work. He writes "8 hours site". He doesn't distinguish between actual installation hours, waiting time because the scaffolding wasn't ready, travel for a missing part.
Eight hours become a single block. And when you compare them with the 8 budgeted, everything looks fine. But those 8 hours aren't all productive. And often they're not 8 but 9 or 10, because the report rounds down. Every time.
The three realities the report hides
When we tracked the real hours, three problems emerged.
Every job ran over by 12-15 hours on average. On a job budgeted at 80-90 hours, 95-105 were always needed. Removing old frames was underestimated, site surprises weren't in the quote, finishing always took longer than planned.
Dead time was worth 4-6 hours per team per week. Waiting for materials. Opening not ready because another tradesman was running late. Trips for forgotten parts. Half an hour here, forty minutes there. In the report it all ends up inside the "8 hours worked".
Overtime wasn't charged to the jobs. Teams were doing 9-10 hours a day to finish on time. Those extra hours were paid on the payslip, but nobody was loading them onto the specific job.
What the lost time costs
| Without tracking | With tracking | |
|---|---|---|
| Dead time per team | 5 h/week | 1.5 h/week |
| Hours lost (2 teams, 46 weeks) | 460 h/year | 138 h/year |
| Cost at 35 €/h | 16.100 € | 4830 € |
| Margin recovered | 11.270 € |
Dead time only. Not counting hour overruns and unbooked overtime.
The system in three steps
Step 1: hours tracking by voice note
The site manager sends a 30-second voice note every time they change activity.
"8:15am, starting removal of old frames, first floor." "10:40am, stopped, second-floor scaffolding not ready." "11:20am, back to work, fitting box frame."
The voice note is transcribed automatically. From it you extract: site, activity, duration. Thirty seconds of voice note is less effort than 10 minutes of a report at the end of the day.
Step 2: weekly comparison
Every Friday, 15 minutes. For each active site: how many hours budgeted, how many used. If a 90-hour site has already used 70 hours and the work is 60% done, there's a problem. You see it on Friday and can act on Monday.
If you wait until the job is finished, you can only note the damage.
Step 3: updating estimation parameters
After three months of real data, redo the parameters used in quotes. Removing aluminium frames from the 1980s takes 40% more hours than recent timber frames. Before, everything was quoted the same. After, you differentiate.
The client price rises accordingly. And you don't lose jobs, because the price was competitive anyway. Just too low.
Track real hours with BAU Gest
BAU Gest compares budgeted and actual hours per job, in real time. You immediately see where margin is running over and act before it's too late.
See how it worksMade
Results after six months
| Before | After | |
|---|---|---|
| Dead time per team | 5 h/week | 1.5 h/week |
| Hour overrun per job | 12-15 hours | 3-4 hours |
| Actual margin | 18% | 26% |
| Overtime | frequent | down 35% |
The reduction in dead time came from two simple interventions: delivering materials the day before (not the same day) and a phone check with the other tradesman before sending the team to site.
The mistake to avoid
"Just buy a timesheet app." No. I've seen companies buy software and after three months nobody was using it. The site manager has dirty hands, work gloves, and no desire to open an app with six menus and twelve fields.
A system only works if using it is easier than not using it. A WhatsApp voice note is easier. A QR code to scan is easier. A digital report with twelve mandatory fields is a burden.
The other trap is collecting data without looking at it. If nobody opens the report on Friday, you've just added bureaucracy.
What to do Monday morning
First: choose the next two jobs and ask the site manager to send a voice note every time they change activity.
Second: at the end of the week, compare actual hours with budgeted hours. Note the gap.
Third: after a month, update the parameters in your next quote. If renovations always run 20% over, add 20% to the estimate.
Hours are money. The money you don't track is the money you lose. If you'd like to understand where margin is evaporating in your business, book 30 minutes with us. Often the numbers from the last 3 months are enough to find the biggest losses.
You might also be interested in
Digital site management: from paper to voice notes in 5 steps
How to move from paper documents and WhatsApp to a digital site management system. A practical guide for those in the construction industry.
Monitoring margins in construction: controlling the numbers
How to calculate and monitor gross, real and net margins in construction, and why small projects are often the biggest problem.
How to delegate in a window and door company without losing control
Are you the man (or woman) who does everything? Here's how to build a delegation system that works in the construction industry, maintaining quality and control.