Which Jobs to Say No To
Most trade businesses that are flat out and still short of money have a filtering problem, not a pricing problem. Here are the signals that a job will lose money before you've even quoted it.
There's a version of a trade business where the phone never stops, the crew is booked six weeks out, and there's still nothing in the account at the end of the month. It almost always comes from the same habit: saying yes to everything that rang. Every yes spends a day of capacity you can't get back, and the jobs that quietly cost you are usually identifiable well before you price them. This guide is a filter — the signals worth acting on, the ones that don't look like problems, and a five-minute test you can run at the ute before you commit.
Doing this job for a living? See job management in ServiceYak.
Busy is not the same as profitable
When you take every job that comes in, your calendar gets allocated by whoever rang first. Not by what pays, not by what you're good at — by timing. Over a year that produces a business shaped entirely by other people's phone habits.
The cost that never shows up anywhere is the one that matters most. A marginal job doesn't just make less money; it occupies the week that the good job would have gone in. You never see that second job, so the loss is invisible, and the only symptom is being exhausted and even.
Before you can filter anything, you need to know which of your finished jobs actually made money. If you've never costed them, that's step zero — finding your break-even point covers the arithmetic, and the pattern in your own job history will surprise you more than any list in an article.
The signals, before you've quoted
None of these are rules. Each one is a reason to slow down, ask a question, or price the risk in properly — and two or three together is usually a no.
| Signal | What it usually turns into |
|---|---|
| The scope changed between the phone call and the site visit | It will keep changing after you've priced it. Every change you don't capture in writing is work you do for free |
| They opened on price | The first question was what it costs, before what it is. Price is the only thing being bought, and there is always someone cheaper |
| You can't get on site to look at it | You'd be pricing a photo and a description. Everything you can't see becomes your risk the moment you send a number |
| It's outside your trade, or at the very edge of it | You're funding your own learning curve on a fixed price, and carrying the warranty on it for years afterwards |
| The travel is a bigger number than the job | Three hours each way for half a day's work spends a full day of capacity to earn half a day's money |
| Two other tradies have already been there | Either the job is harder than it looks or the client is. Both are worth knowing before you quote, and one phone call usually tells you |
| They want you to start before anything's in writing | The urgency is the tell. A written scope and a deposit resolve it very quickly, one way or the other |
| The person you're talking to isn't the person who pays | A landlord, an insurer, a strata committee, a head contractor, a partner who hasn't seen the price. Find out who signs and who pays before you start |
The ones that don't look like a problem
The obvious bad jobs are easy. These are the ones that cost established businesses real money, because every one of them arrives looking like good news:
- The big job that fills the calendar. One client at 40% of your turnover isn't a customer, they're an employer — and their payment terms have just become your cash flow. Concentration is the risk, and it's invisible until the month they pay late.
- The good client asking for work you don't do. The hardest no in the trade, because you don't want to lose them. Refer it. Attempting it is how you turn a decade-long relationship into a warranty claim.
- The one you take to keep the crew busy. Sometimes exactly right. Often it's the job that stops you selling properly for a fortnight and locks in a rate you'll be held to next time.
- The small, urgent, after-four regular. Each one is fine. Forty of them a year is a business that can't schedule anything, priced as though the disruption were free.
- Mate's rates. Never one job, never fully scoped, and the hardest place in the world to raise a variation. Decide what you're gifting and treat the rest like work.
- The job in a sector you don't know. New build, insurance, strata, government and defence work all have their own paperwork, payment cycles and retention. The trade is the easy part.
Look at where your own bad debts have come from over the last few years. For most trade businesses they cluster — one client type, one kind of work, one part of the market. That cluster is the most valuable filter you'll ever build, and it's already sitting in your invoice history.
The five-minute test
You don't need a spreadsheet for this. Six questions, answered honestly before you commit to quoting:
- What's it worth per day of capacity?. Not the total. A $12,000 job that runs ten days is $1,200 a day; a $3,000 job that runs one is $3,000. The big number on the quote tells you nothing on its own.
- Who pays, when, and what if they don't?. Deposit, terms, and whether this payer has ever paid you before. If the answer to any of those is uncomfortable, the price needs to reflect it or the job isn't worth having.
- Can I see the whole job?. If not, is it genuinely estimate-able with stated assumptions, or are you writing a blank cheque to yourself? Quote vs estimate is the difference.
- Is it work I do properly?. At a standard you'd put your name on, with a warranty you're happy to carry for the next several years. Any hesitation here should be a no, regardless of the money.
- What does it cost me just to be there?. Travel, parking, site access, inductions, permits, waiting for other trades. On some jobs this is the entire margin and it never appears on the quote.
- What am I turning down to take it?. The one people skip. If your next six weeks are full, this job has to beat the work you'd otherwise take, not just cover its own costs.
If three of the six come back badly, it isn't a pricing problem you can fix with a bigger number. Occasionally it's worth quoting anyway at a price that makes the risk worth carrying — but quote it to be paid properly, not to win it.
Saying no without burning it
Trade work runs on word of mouth, and a badly handled no costs you referrals from people who were never going to be clients themselves. It takes about a minute to do properly.
- Be quick. The worst no is the one delivered by not sending a quote and not returning the call. That's the version people tell other people about.
- Say it's not your work, not that they're not your client. "That's not really what we do" or "we don't cover that far south" is honest and impersonal.
- Refer someone genuinely suited. It costs you nothing, it makes you the person they ring next time, and referrals come back the other way.
- Don't explain at length. A paragraph of justification reads as an opening position, and you'll be negotiating five minutes later.
- Or price it properly and mean it. Sometimes they say yes at your real number, and then it's a good job rather than a no.
The filter is also a plan
Saying no on its own just makes you quieter. The point is what you do with the space: once you know which work actually makes you money, you can go and get more of that specific thing rather than more of everything.
- Narrow what you advertise to the two or three services you want more of. A tradie who says he does everything gets asked about everything.
- State your service area honestly, so the three-hour jobs stop arriving in the first place.
- Put a minimum charge in front of people if small call-outs are eating your week.
- Turn the good work into repeat work — the maintenance client you already know is worth more than a new enquiry you don't.
How to get more work as a tradie covers the channels for that. Then revisit the filter each year — the rules that were right for a one-van operation are usually wrong at three, and the jobs you had to turn down at three vans are often the ones worth building around.
ServiceYak costs every finished job against what you quoted, so the filter gets built from what your work actually made rather than from which jobs you remember being painful. The two lists are rarely the same.
Frequently asked questions
How do I know if a job is worth taking?
Work out what it pays per day of capacity rather than in total, check who pays and on what terms, confirm you can actually see the whole job, be honest about whether it's your trade, and count what it costs simply to be on site. Then ask what you're turning down to take it — that last one decides more than the price does.
Should I turn work down when I'm quiet?
Sometimes, but not out of principle. A thin job that covers wages beats an empty week. What a quiet week is never a good reason for is accepting a scope that won't hold still, a client who won't pay a deposit, or work outside your trade — those cost more than the gap in the calendar would have.
How do I say no to a good client?
Quickly, and about the work rather than about them. "That's not really what we do, but I know someone who does it properly" keeps the relationship intact and usually gets you the referral back. Attempting work outside your trade to avoid disappointing someone is how good relationships end.
Which jobs most often lose money?
Honestly, whichever ones you haven't costed. The recurring patterns are small unscoped call-outs with long travel, work where you couldn't see what you were taking on before you priced it, and any single client big enough that their payment terms set your cash flow.