Why Jobs Blow Out, and How to Catch It Early
A blown job is rarely one bad day — it's four small leaks nobody watched. Here's where the hours and materials actually go, the signals that show up while you can still act, and what to do the moment you spot one.
Every tradie knows the feeling of finishing a job that went fine, invoicing it, and quietly wondering where the money went. By then it's history — you can learn from it, but you can't fix it. The useful question is a different one: how do you tell, while a job is still running, that it's heading over? This guide covers where blow-outs actually come from, why so many of them start before the crew turns up, the specific signals that show up early, and what to do about each one while there's still something you can do.
Doing this job for a living? See job management in ServiceYak.
It's never one bad day
Blow-outs don't usually announce themselves. There's no moment where something goes obviously wrong — just a couple of extra hours on Tuesday, a second run to the supplier, a bit the client asked for that nobody wrote down, and a job that drifts into a day it wasn't meant to need.
Any one of those is nothing. Together they're your margin, and because none of them felt like a problem at the time, nobody mentions them. You find out at invoicing, which is the one point in the job where you can't do anything about it.
Working out what a finished job made is job costing, and it's how you price the next one better. This is a different exercise: watching the job you're standing in, while there are still decisions left to make. Both are worth doing — but only one of them can save the job that's running.
Most blow-outs start before anyone picks up a tool
When a job runs over, the easy read is that the crew were slow. Sometimes that's fair. Far more often the first morning was set up so that the first half-day was never going to be productive, and no amount of effort recovers it.
The version everyone recognises: your tradie arrives, spends the best part of an hour working out what the job actually is, drives off for materials because nothing was ordered, stops for a coffee on the way back, and starts work near enough to lunchtime. You've paid four hours of wages and you can charge for none of them. The job hasn't blown out yet, but it already has.
- Materials weren't ordered or weren't on site, so the first move of the day is a drive.
- Nobody was told the scope, so the crew reconstruct the job from the plans and their best guess.
- Nobody was told the hours. The quote allowed thirty; the person doing the work has no idea whether that's generous or tight.
- Access wasn't sorted — keys, parking, a site contact, another trade still in the room.
- The job was never properly scoped, because it was priced off a quick look and a phone photo.
- Exclusions weren't passed on, so the crew do the extra bit to be helpful and nobody bills it.
The counter to all six is the same fifteen minutes. Meet the crew on site at the start, walk it together, say what's been quoted, what's excluded, how many hours are in it and when it needs to be done. Then leave. That quarter of an hour is the highest-return time you will spend on the job, and it's the one most often skipped because you've got somewhere else to be.
The four places a job leaks
Nearly every blow-out is one of four things, or a combination. Naming which one you're looking at is the first step, because the fix is completely different in each case.
| The leak | What it looks like on site | Whether it's recoverable |
|---|---|---|
| Labour running over | The work is simply taking longer than the hours you allowed | Not from this client — but often fixable mid-job if the cause is a blocker rather than pace |
| Non-chargeable time | Travel, supplier runs, waiting on another trade, standing around for a decision, return visits | Rarely billable, almost always preventable with better setup |
| Materials | Extra trips, wastage, the wrong item, a price that moved since you quoted | Sometimes — a genuine price rise on a long-dated quote can be a fair conversation |
| Unpriced variations | The client asked for something, someone said yes, nothing was written down | Fully recoverable, but only if you catch it the same day |
| Rework | Going back to fix something, or redoing work after another trade | Depends entirely on whose fault it was, which is why photos matter |
The fourth row is the one worth the most attention, because it's the only leak where the money is genuinely still there. An extra you catch on the day is a variation you can price and get agreed. The same extra found at invoicing is an argument you'll probably lose.
The signals that show up while you can still act
You don't need live figures on a screen to spot a job going off. You need a few checks you do consistently, and a threshold that turns a feeling into an action.
- Hours burnt against work done. The single most useful check there is. At half the allowed hours, is half the job finished? If you allowed thirty and you're twenty in with a third to go, you already know how this ends.
- The materials figure moving before the job is half built. If most of the materials budget is gone and the structure isn't up, something was underpriced or something is being wasted.
- A day quietly appearing on the schedule. "We'll just need Friday to finish off" is a blow-out being announced in the politest possible terms.
- A crew member who's gone quiet, or won't give you a finish date when you ask for one. People don't hide good news.
- Something on site that isn't what you quoted — a different make, an extra circuit, a wall that wasn't in the scope.
- Supplier dockets on the job you don't recognise. Every unexpected docket is a question worth asking the same day.
- The client mentioning things in conversation rather than in writing. "While you're here" is the most expensive phrase in the trade.
Set a threshold rather than trusting your gut. Something like: check every live job at a quarter and at half of its allowed hours, and anything more than ten per cent off the line at the halfway mark gets dealt with that week. A number decides for you on the day you're too busy to think about it.
What to do the moment you spot one
The instinct is to ring the crew and ask what's going on. Work out which of the four leaks it is first — the response is different for each, and the wrong response makes it worse.
- If it's an unpriced variation, stop and price it. Before another hour goes into it. Write up the extra scope and the number, send it, and get an explicit yes. This is the one category where the money is still fully available to you, and it stops being available the day the job finishes.
- If the crew's run rate is off, go to site. Don't do it on the phone. Nine times out of ten it isn't effort — it's a decision nobody has made, a part that hasn't arrived, another trade in the way, or a detail on the plans that doesn't work. All of those are yours to clear, and they clear fastest in person.
- If it's materials, find out which kind. Wastage, extra trips and wrong orders are process problems for next time. A genuine supplier price rise on a quote you sent months ago is a conversation you're entitled to have with the client — which is exactly what a validity period on your quote is for.
- If it's your quote, take it on the chin and write down why. You can't go back to the client because you priced it short. Finish it properly, protect the relationship, and record the reason against the job while it's fresh. That note is worth real money the next time you price similar work.
- Tell the client early if the date is moving. A date you flag a week out is a rescheduling. The same date missed in silence is a complaint, and it turns a job that lost a bit of margin into one that also costs you a referral.
- Re-forecast the rest of the job. Don't just note the overrun — work out what the remaining work now needs. Half the value of catching it early is deciding what to do differently over the days that are left.
Make the check part of the week
None of the above happens by good intentions. It happens because there's twenty minutes in the week with your name on it, at the same time each week, where you go through the live jobs.
Three questions per job is enough:
- Hours used against hours allowed — and is the work at the same point?
- Has anything been bought for this job that wasn't in the quote?
- Has anything been agreed on site that hasn't been priced?
Then log the reason on every job that ends up over. After a quarter of doing that you'll have your own list rather than a general sense of being flat out and not far enough ahead — and most trade businesses find that two or three causes account for nearly all of it. They're usually not the ones the owner would have guessed.
ServiceYak keeps the quote, the hours, the materials, the photos and the variations on the one job record as they happen — so a job that's drifting shows up while you can still do something about it, rather than at invoicing when all that's left is the arithmetic.
Frequently asked questions
How do I tell a job is going over before it's finished?
Compare hours used against hours allowed, and check whether the work is at the same point. At half the allowed hours, roughly half the job should be done. Alongside that, watch for materials being bought that weren't quoted and for anything agreed on site that hasn't been priced — those three checks catch most blow-outs while you can still act.
Whose fault is it when a job blows out?
More often the setup than the crew. A job where nobody was told the scope or the allowed hours, and where materials weren't on site on the first morning, loses half a day before anyone picks up a tool. Look at what the crew were given before you look at what they did with it.
Should I tell the client a job is running over?
Tell them about anything that affects them — the date, or extra work that needs pricing — and tell them early, because a flagged delay is a reschedule and a silent one is a complaint. You don't need to tell them your margin is thin. If the overrun is down to your own quote, that's yours to wear.
What do I do about a job I've already lost money on?
Finish it properly. Cutting corners on a job that's already gone costs you a referral on top of the margin. Then write down which of the four leaks it was — labour, non-chargeable time, materials or an unpriced variation — while you still remember. That note is the only thing a bad job gives you, and it's worth having.