Follow one job through your own office

Crew22 Pty Ltd Published 12 September 2026

Most contracting businesses have never seen their own paperwork drawn out. Not because nobody cares. Because the job that pays this week always beats the job that explains the year, and everybody in the office already knows the system is clunky, so there seems little point proving it.

There is a point, and it is a number. Once you can say out loud how many hours a year your office spends typing the same job in twice, you can decide whether that is worth fixing. Until then it is a feeling, and nobody funds a feeling.

This is the walk we do on the first day of an Audit. Written out here so you can run it yourself. It takes about an hour, you need one finished job and a pen, and the answer belongs to you whether you ever talk to us or not.

What you are actually looking for

You are not looking for slow people. You are looking for re-entry: the same fact, typed by a human being into a second screen because the first screen cannot pass it on.

A job number. A plant ID. A start time. Seven and a half hours on a 22 tonne excavator. Every one of those facts was captured correctly the first time, on site, by somebody who was there. Everything that happens to it afterwards is copying. Copying is where the hours go, and it is also where the mistakes come from, because a number typed four times will eventually be typed wrong once.

What you need before you start

• One finished job. Already delivered, already invoiced, already paid if possible. A job still running will keep changing under you.
• A middling one. Not your biggest and not a half day. Pick the size of job you do most weeks, because that is the one the year is made of.
• An hour, and the people who touched it. The person who took the docket, the person who did the timesheets, the person who raised the invoice. Often two of those are the same person, which is worth noticing on its own.
• Paper. Do this on paper the first time. A spreadsheet makes you tidy it up while you are still finding it out.

Step one: start where the fact was born

Go to the very first place the job existed as writing. Usually that is a docket, a start slip, a text message from a supervisor, or a scribble in a diary. Do not start at the office. If you start at the office you will miss the handover that costs you the most.

Write the date, the job number, and the words first written here. That is line one.

Step two: write a line for every screen it lands in

Now walk it forward. Every single time that job appears somewhere new, write another line. Four columns, no more than that:

Keep going until you reach the invoice going out the door, and then keep going one more step, because the last one is usually the one nobody mentions. Somebody reconciles it. Somebody files it. Somebody sends the docket photo to the client because the client asked.

A normal run for one earthmoving job looks something like this, and yours will be longer than you expect.

1. Docket written on site.
2. Photo of the docket sent to the office.
3. Hours keyed into a timesheet.
4. Hours keyed again into the job costing.
5. Plant hours keyed into the maintenance sheet.
6. Invoice line raised in the accounting package.
7. Docket attached to the invoice by hand so the client will approve it.

Seven places, one set of numbers, and six of them are copying. Nobody designed that. It grew, one sensible decision at a time, over years.

Step three: mark the copies

Go back through and highlight every line where the information already existed somewhere else in the business. Those are your candidates. Leave the first capture alone, because that one is real work, and leave anything where a human genuinely made a decision, because that is judgement and it is the part you are paying for.

What is left highlighted is the pile. It is normally bigger than anyone in the room guessed, and there is usually one line that makes somebody laugh.

Step four: turn it into a year

Add up the highlighted minutes for that one job. Then:

Forty eight weeks is our assumption, not a law. It allows for shutdown, public holidays and the weeks when nothing goes to plan. Use your own number if you have one. The point is not decimal places, it is the order of magnitude. There is a big difference between deciding about forty hours a year and deciding about eight hundred, and most offices do not know which one they are looking at until they do this.

If you would rather move three sliders than do the arithmetic, the hours tool on this site does exactly this sum and shows its working. It asks for nothing and sends nothing unless you press the button.

Step five: ask the better question

Here is where most people get this wrong, so it is worth saying plainly. The answer to "we spend 600 hours a year copying" is not "we need fewer people". It never has been. Those 600 hours are not a saving, they are capacity that is already on the payroll and currently being spent on typing.

The useful question is the other one: what would this office get done if it had those hours back? Usually the honest answer is quotes. Quotes going out the same week instead of a fortnight later. Sometimes it is invoicing sooner, which is money in the bank earlier for work already done. Sometimes it is one person finally getting to the thing they have been promising to get to since March.

Write that answer down next to the number. The pair of them together is a business case. The number on its own is just a complaint with a decimal point.

Step six: pick one, and only one

Look at your highlighted lines and choose the single one with the best ratio of hours saved to difficulty. It is almost never the biggest problem. It is usually a boring one, and boring is good, because a boring fix that works this month beats a grand plan that lands next year.

Two rules for picking it. It has to be a fact that already exists in a system, not one that lives only in somebody's head or on a photo of a docket. And it has to be something the crew does not have to relearn, because a fix that changes how site works will quietly not get used.

Sometimes the answer is that you are fine

It happens. Some businesses are tidier than they think, usually because one person years ago set something up properly and nobody ever broke it. If you run this and the highlighted pile is small, that is a genuinely good outcome and it cost you an hour to find out.

You will still come out of it holding a drawing of how your own business actually runs, which most operators have never seen, and which is worth having the next time someone tries to sell you a platform that replaces everything.

What we do differently, in one paragraph

When we do this we do it with a stopwatch and an outsider's eyes, we do it across several jobs rather than one, and we come back with a single page: how a job really moves, every place it gets typed twice, what those hours add up to across a year, and which one to do first. That is the Audit, it costs $2,500, and it comes off the cost of the work if you go ahead. If we find nothing worth fixing we tell you that and send you the page anyway.

But you do not need us to start. You need one finished job, an hour, and the willingness to count something you already suspected. Run it this week.