Performance Reviews in a Trade Business
Twenty minutes, twice a year, built on what the job records already show. Here's what to review in a trade business, where the evidence comes from, and how to run it without turning into a corporation.
Most trade businesses don't do performance reviews, and the reason is fair enough: the ones people have experienced were corporate rituals with rating scales and forms nobody read. That's not what this is. In a business of four or six people, a review is twenty minutes twice a year where you both say the things that otherwise never get said — what's going well, what needs to change, what they're working towards, and when pay gets looked at. It's cheap, it prevents most of the problems in the guides either side of this one, and it works best when it's short.
Doing this job for a living? See job management in ServiceYak.
What a review is actually for
Not paperwork, and not a stick. In a small trade business a review does four jobs that nothing else does:
- It gets the small things said before they've been irritating you for eight months and come out badly.
- It gives progression a shape. "What do you want to be doing in two years" is not a question that comes up on its own in a ute at 6am.
- It fixes the pay conversation to a date, so nobody has to work up the courage to ask and you're not making a decision on the spot in a car park.
- It creates a record — dated, agreed, both directions. If a situation ever does go bad, a history of honest reviews is worth more than anything you write afterwards.
Twice a year is plenty. Once a year is workable. Quarterly, in a business this size, becomes a chore that gets skipped by the third one.
What to review in a trade business
Skip the generic competency lists. These are the areas that actually determine whether someone is good in a trade business, and they're all observable:
| Area | What you're actually looking at |
|---|---|
| Quality of finish | Callbacks, rework, defects picked up at handover, and whether the standard holds when nobody's watching |
| Pace against the quote | How their time on jobs compares to what was quoted — over several jobs, not one bad one |
| Safety | PPE, method, whether they raise hazards, whether the paperwork gets done before the work starts |
| Job records | Photos, notes, materials and variations captured on site — the difference between a clean invoice and an argument |
| Clients | How they are in someone's house, what the reviews say, whether clients ask for them by name |
| The crew | Whether they help an apprentice, whether they're a person others want on their job |
| Independence | The job types they can now run without ringing you, compared with six months ago |
| Tickets and licences | What's current, what's expiring, what they need next for the work you're moving into |
Weight it towards the things you can point at. "Your finish has been consistently good and there've been no callbacks since March" lands. "You've got a great attitude" is nice to hear and teaches nothing.
Where the evidence comes from
The rule that makes reviews work is that nothing in one should be a surprise. That means the review is a summary of things already noticed, and it means you need a source other than memory — which reliably over-weights the last fortnight.
- Job records — who did what, how long it took, what was captured on site. The most objective thing you have.
- Quoted versus actual hours across a run of jobs. Read it carefully: a consistent gap on one person's jobs might be them, or it might be that they always get the ugly ones.
- Callbacks and defects, logged as they happen with the job attached.
- Client feedback, good and bad, including the online reviews that mention someone by name.
- A running note. Thirty seconds when something happens — a rescued job, a complaint, a good call on site. Without it you'll write the review from the last three weeks.
Be careful with numbers on their own. Hours against quote tell you something real, but a person who's been carrying the hardest jobs and training the apprentice will look worse on that measure than someone doing straightforward repeat work. Use the figures to start the conversation, not to end it.
Running the conversation
- Give them notice and a couple of questions. A week ahead: what's gone well, what's frustrating, what do you want to be doing in two years. People who've thought about it turn up with something to say.
- Pick somewhere you won't be interrupted. The office, a cafe, or the ute parked up. Twenty to thirty minutes, phones down. Not on site with a client waiting.
- Let them go first. How do you think the last six months have gone. Most people are harder on themselves than you'd be, and it tells you whether you're both looking at the same picture.
- Give two things to keep and two to change. Specific and dated. More than two on each side and none of it is remembered.
- Agree what happens next. A ticket to get, a job type to start running alone, a standard to lift. Two or three at most, with dates against them.
- Ask about the business. What's getting in the way, what would you change if it were yours. This is where you find out about the tool that's been broken for a month.
- Write it up the same day. Half a page: what was discussed, what was agreed, when you'll check in. Send it to them. Diarise the follow-up, because an agreed goal nobody revisits is worse than none.
Where pay fits
This is the part people get tangled in. Two workable approaches, and the wrong move is having no position at all:
- Separate the conversations. Review performance in one meeting, pay a few weeks later. The advantage is that the feedback actually gets heard instead of everyone waiting for the number. The risk is looking like you're dodging it — so say up front when the pay conversation is happening.
- Or do it together, and be straight. If it's a yes, say what it's for. If it's a no, say why and say when you'll revisit it.
Either way, don't let a review pass without acknowledging that pay exists. Silence gets read as a permanent no, and that's the read that puts someone on the market.
Award classifications, minimum rates and allowances are set externally and change over time. Check the current position through Fair Work for the award that covers your trade before you commit to anything in a review.
The ways reviews go wrong
- Saving up a live problem for the review. If something's wrong in March, deal with it in March. The review tracks whether it stuck.
- All praise, no substance — pleasant, useless, and it makes the eventual hard conversation look like it came from nowhere.
- Recency bias. A great year and a bad fortnight becomes a bad review unless you've kept notes.
- Goals with no date and no follow-up. Nothing teaches people to ignore reviews faster.
- Skipping your best performer because there's nothing to fix. That's the person most likely to be recruited away and least likely to tell you they're unhappy.
- Copying a corporate template. Rating scales out of five in a business of five people don't measure anything you don't already know.
ServiceYak gives you the factual half of a review without any extra admin: every job is assigned to a person, time is logged against it, and photos, notes and variations are captured on site — so you can look back over six months of real work instead of writing the review from what you remember of last week.
Frequently asked questions
How often should I do performance reviews in a small trade business?
Twice a year is the sweet spot — often enough that nothing festers, rare enough that it stays a real conversation rather than a chore. Once a year is workable if you're also having short informal check-ins. Quarterly reviews in a business of a handful of people usually get abandoned by the third round.
What should I actually measure?
Quality and callbacks, time against quoted hours across a run of jobs, safety, how well job records come back, how they are with clients and the crew, and what they can now run without you. All of it should come from things you can point at, not impressions formed in the last fortnight.
Should pay be part of the review?
Either separate them or do them together, but decide and say which. Separating means the feedback gets heard rather than everyone waiting for the number; combining is simpler. What doesn't work is a review that never mentions pay, because silence gets read as a permanent no.
Do I need to keep records of performance reviews?
Keep them regardless. Half a page, dated, sent to the person, is enough. It's useful for tracking progression and for the pay conversation, and if a situation ever turns into a formal performance issue, a history of honest, dated reviews is worth far more than notes written after the fact. Fair Work publishes what employment records have to be kept.