The job went well. The customer is happy, the crew finished on time, the invoice went out. Three months later you realize that contract made you almost nothing. Nobody stole anything: there are just hours missing. Two guys who stayed an extra hour on Tuesday, a run to the supply house nobody wrote down, a half-day of rework everyone forgot. All of them got paid. None of them got billed.

Short answer: your timesheet has to capture four things (who, when, which job site, which task), get filled in the same day, get approved, and be connected to your project costs. This guide covers what badly recorded hours cost you, the tracking methods and when each one breaks, GPS time tracking and its limits, and how to know whether a job was profitable.

What badly recorded hours really cost you

A badly recorded hour makes no noise. Nobody complains, and that's exactly why it's expensive. It drains you in four ways.

The hours that vanish

The employee wraps up at 4:40 instead of 4:00, drives back for a material, waits thirty minutes on a delivery. On Friday he fills in the sheet from memory and writes 8 hours, because that looks like a normal day. Those bits of time leave your bank account. They never make it onto your invoice.

The underbilled invoice

On time and material work, an unrecorded hour is an unbillable hour. On a fixed-price job it's sneakier: the amount doesn't move, so you see nothing. You find out at the end that the job cost more than planned, without knowing where.

The dispute with the customer

The customer challenges twelve hours of extras. What do you say? If your only evidence is a sheet filled in from memory and your word, you'll split the difference. With hours recorded the same day, by employee and by task, you have something to answer with, line by line.

Payroll by approximation

Fuzzy hours make fuzzy payroll. Overpaid, it costs you directly. Underpaid, it costs you more: a good worker chasing his own hours two pay periods in a row starts listening to other offers.

Paper, texts, spreadsheet or app: what each method costs you

There's no shameful method. There are methods that hold up to a certain volume, then give out. And the real problem is almost never collecting the hours: it's compiling them.

• Paper. It works anywhere, no signal, no battery, and everyone knows how to use it. The cost is elsewhere: someone retypes it all by hand and sheets get lost in a truck.
• Texts. Fast, adopted with zero training, but everything sits loose in a thread: no totals, nothing tied to a job site, and finding last Tuesday is an archaeology project.
• The spreadsheet. Plenty of profitable contractors run this way, and that's not a flaw: it calculates, it exports, it costs nothing. Its limits show up with volume: manual entry and formulas a copy-paste breaks.
• A time tracking app. The employee clocks in on his phone, hours attach to the right job site, compiling happens on its own. The cost is real: a subscription and an adjustment period.

The tipping point isn't a matter of taste, it's a matter of volume. With two or three employees on one site, paper or a spreadsheet hold up fine. The moment you're running several crews across several sites in the same week, or compiling takes you an evening, you're already paying for an app. You're just paying in office hours.

The four things your timesheet has to capture

1. Who. The employee, by name. Not "the crew," not "the guys."
2. When. Start time, breaks and end time, not just a daily total. An "8 hours" written on Friday isn't data, it's a memory.
3. Which job site. The specific project or customer. An hour attached to nothing is useful to nobody.
4. Which task. Demolition, install, finishing, rework, waiting. This is the field everyone skips.

That fourth field is the most valuable one. Without it, you know a job took 240 hours but not why. With it, you find out where the time actually goes, like finishing running longer than you quote. Two jobs priced the same, one profitable and one not: the task field is what tells you which is which. Three add-ons also pay for themselves:

• Travel time, recorded separately from work time. Decide your rule once (paid or not, starting from where) and apply it everywhere.
• Materials pulled off the truck, entered while you're at it. That's the other half of your job cost.
• A note or a photo when something goes sideways: a delay, blocked access, rework caused by another trade.

When to fill it in: same day, not Friday

The Friday night catch-up is the easiest way to get the hours wrong, and it isn't a question of honesty. Nobody remembers Tuesday precisely. You round, you smooth, you put 8 hours everywhere because it looks about right. A sheet filled in from memory is an estimate wearing the name of data.

1. The employee records his hours the same day, at the end of his shift.
2. You or the foreman approve the week's hours before they go to payroll. Five minutes that prevent a retroactive correction.
3. Any later change leaves a trail: who changed what, when, and why.

That last point isn't bureaucracy. When someone forgets to clock in, you need a clear correction procedure, not an eraser. An hour corrected with a written reason stays defensible in front of a customer or an auditor. An hour changed quietly does not.

GPS time tracking: what it fixes and what it doesn't

A mobile time clock fixes three real problems: the time is captured as it happens, the job site is picked from a list, and retyping disappears. Several tools on the market add geolocation to confirm the punch happened on site. Before you turn that on for your crew, two things deserve an honest look.

Privacy isn't a detail

In Quebec, monitoring an employee isn't forbidden, but it isn't a free-for-all either. The general principle: you need a serious reason, a measure that is targeted and proportionate, the least intrusive means available, and employees have to be informed when a technology can identify or locate them. The framework comes from the Charter of Human Rights and Freedoms and Quebec's private-sector personal information protection law. On the ground: capturing a position at the moment of the punch is not the same thing as tracking someone all day, and the second one is considerably harder to justify. These are general principles, not legal advice: your specific situation gets validated with the Commission d'accès à l'information or with legal counsel.

Getting the crew on board

A point few contractors anticipate: an employee's consent is never fully free when refusing puts his job on the line. And a punch someone forgets to close keeps capturing locations long after the job site. A tool that fixes a billing problem can create a workplace one.

• Explain before you install, never after. An employee who discovers geolocation on his phone will assume the worst.
• Put in writing what gets captured, when, why, and who can see it.
• Capture at the punch only, never continuously, and give every employee access to his own data.
• Ask yourself whether you have a location problem or just a compiling problem. Most of the time it's the second one, and it gets solved without geolocation.

Linking hours to costs: knowing whether a job was profitable

This is where time tracking stops being paperwork. An hour recorded on the right job site, with the right task, becomes a line in your cost of doing the work. Without it, you don't know whether you made money: you only know the bank account isn't empty.

The basic calculation comes down to one idea: the true cost of an hour is not the employee's hourly rate. It's the wage, plus the charges that come with it, plus the equipment and the vehicle that hour consumes. Your exact rates depend on your sector and your collective agreement: pull them from your financial statements, not from a blog post.

1. Establish your loaded hourly cost by type of worker, from your real numbers.
2. Attach every hour to the right job site and the right task.
3. Compare two columns: hours estimated in the quote and actual hours to date.
4. Look at the gap during the job, not three months after the final invoice.
5. Carry the finished job's actual hours into your next quote for the same type of work.

An hour nobody recorded didn't just disappear from the invoice. It disappeared from what you know about your own business.

The second-to-last step is the one that saves contracts. A fifteen-hour gap spotted mid-job can be recovered: you adjust the sequence, you discuss an extra while the customer still has the work in front of him. Discovered while preparing the final invoice, it's just bad news. And the last step closes the loop: your finished jobs become your best pricing tool.

Construction: the CCQ reality

If you fall under Quebec's construction act (R-20), part of this work isn't optional. The CCQ requires a register by employee and by job site, plus periodic reporting. The exact content and the form of that register are set by regulation: pull the current list straight from the CCQ before you build your template.

The good news is that the work doesn't double up. A timesheet that already captures the four things above gives you the raw material, both for your register and for your real cost. You're starting from a template to complete, not from a blank page.

One clear warning though: the exact rules, deadlines, retention periods and rates vary by sector, by collective agreement and by your situation, and they change. Verify them with the CCQ, and with the CNESST for labour standards outside construction. A blog, this one included, is not a regulatory source.

In ORYX, the Time clock does exactly this work: your employees clock their hours from their phone on site, timesheets compile on their own and hours stay attached to the right customer and the right project. The punch captures the time and the job site, which solves the compiling problem.

Because hours live in the same place as your projects, quotes and invoices, you see the link between time spent and money earned while the job is still running. Data hosted in Canada, Law 25 compliant.