Guide

The job is running over: when to eat it and when to charge

Set a written variance rule before you're standing in the client's kitchen arguing about hours. Here's the threshold, the wording that makes extra billable, and the maths on what happens if you don't do this.

The moment nobody plans for

You quoted 16 hours labour. You're at hour 20 and the job isn't done. Do you finish it and eat the extra four hours, or do you call the client? Most tradespeople decide this in the moment, tired, halfway through a wall, and get it wrong in both directions: they either swallow overruns that quietly wreck their margin, or they nickel-and-dime a client over noise that was always going to happen.

The fix isn't a better attitude. It's a written rule you set before the job starts, so the decision is already made when the overrun happens.

The three-band rule

Set this out in your terms, or at minimum in your head, before you quote:

• Under 10% over quoted labour hours: absorb it. This is estimating noise. Every quote has a margin of error — parking, an awkward stud, a slow supplier counter. If you priced the job properly you already built some of this in.
• 10–25% over: raise a change order. This is where you stop, tell the client what's happened, and get written agreement before continuing. It might still be your estimating error rather than their scope change — but past 10% it's material enough to disclose, not absorb silently.
• Over 25%: stop work until approved in writing. No exceptions. If a job is a quarter over budget and still not finished, something is structurally wrong with the quote or the scope, and continuing on trust is how small businesses go under.

These are labour-hour thresholds, not job-price thresholds. If you quoted 20 hours and you're at 23, that's 15% — change order territory, even if materials came in cheap and the invoice total looks fine.

Scope creep vs underestimating: they are not the same problem

This distinction decides whether you can bill for it.

Scope creep is when the client adds or changes something after you quoted: "while you're here, can you also do the ensuite tap", a different tile chosen mid-job, access that wasn't available on the walkthrough. This is billable. You quoted for a defined scope; they changed it.

Underestimating is when the job is exactly what you quoted, but it took longer than you priced. Wrong labour hours, wrong material waste factor, didn't account for a second trip. This is not billable to the client — it's your pricing error, and the client shouldn't pay for it. It should feed back into how you quote next time (more on that below).

Mixing these up is what damages trust. Charging for your own bad estimate looks like padding. Not charging for genuine scope change trains clients to keep adding "just one more thing" for free.

The wording that makes extra time billable

When it's genuine scope creep, say this, in writing, before you do the extra work:

"That wasn't part of the original quote for [scope]. I can do it, but it'll add approximately [X hours / £Y]. Confirm and I'll get it done today."

Three things make this stick:

• You name the original scope, so it's clear what's outside it.
• You give a number before starting, not after.
• You get a yes in a text or email, not a verbal nod on site. A verbal "yeah go ahead" is worthless when the invoice arrives.

Keep a standard change-order template — even a one-line text format — so this takes ten seconds, not a negotiation.

What an unmanaged overrun actually costs you

This is the part most trades underprice in their head. A 25% labour overrun doesn't just mean "a bit less profit" — it can gut a job's margin entirely.

Take a job quoted at £2,000 with a 45% gross margin (£900 profit, £1,100 cost — mostly labour). If the labour portion runs 25% over and you absorb it without charging, your cost rises but your price doesn't. Depending on how much of that £1,100 was labour, a 25% labour overrun can push total cost up by roughly £180–£250, dropping your margin from 45% down to somewhere around 25–28%.

Scenario	Quoted cost	Actual cost	Price	Margin
Quoted as planned	£1,100	£1,100	£2,000	45%
25% labour overrun, absorbed	£1,100	~£1,300	£2,000	~35%
Same overrun, on a labour-heavy job	£1,100	~£1,450	£2,000	~27%

Now scale it up. If your shop runs 10 jobs a month averaging £2,000 each at a healthy margin, and just two of those a month run 25% over and get absorbed rather than change-ordered, you can lose the equivalent of £3,000–£5,000 a year in margin that simply evaporates — often enough to erase a small operation's entire annual profit. This is not a rounding error. It's the difference between a business that pays you and one that just pays your bills.

Run your own numbers through the job profitability calculator to see what a specific overrun does to a specific job, and check your baseline margins with the markup and margin calculator before you're relying on guesswork on site.

Worked example: the kitchen re-tile

You quote a kitchen splashback re-tile at 12 hours labour, priced off your standard hourly rate. On the day:

• Hours 1–13 (8% over): old adhesive is tougher to strip than expected. Under 10%. You absorb it, finish the job, say nothing, move on.
• The client then asks you to also tile behind the hob extension, which wasn't in scope. That's 3 extra hours — scope creep. You quote it on the spot in writing: "That's outside the original job, will add about 3 hours / £135, confirm and I'll do it now." They confirm by text.
• Separately, if the original 12-hour job itself had crept to 16 hours (33% over) with no scope change, that crosses your 25% threshold. You stop, explain what's happened and why, and get written sign-off before continuing — even if it means coming back tomorrow.

Same job, three different responses, because the causes were different.

The 90-day feedback loop

Thresholds handle the moment. This loop fixes the pattern, so you're not absorbing the same estimating error every month.

1. Log it. For every job over the next 90 days, record quoted hours vs actual hours. A notebook or a spreadsheet column is enough — it doesn't need software.
2. Compute the average overrun across similar job types (e.g. bathroom re-fits separately from small repairs — don't blend them).
3. Add it back into your quote, plus a 10% contingency on top. If re-fits are averaging 14% over, your new baseline quote for that job type should include that 14%, plus another 10% buffer — not because you're padding, but because your last quote was demonstrably wrong.

After 90 days you'll either have tighter quotes that stop running over, or clear evidence a job type is structurally underpriced and needs a rate review, not just a bigger buffer. Either way, you stop finding out the hard way, one absorbed afternoon at a time.

If you want a repeatable way to track this across a small team — quote sheets, change-order templates, job logs — the small business operations kit has ready-made versions so you're not building spreadsheets from scratch.