Pick it for how it connects, not for what it does.

Why we won’t rank the three

We have implemented against all three — Xero, MYOB and QuickBooks Online — and that is exactly why there is no winner at the top of this page.

A ranking would have to rest on measured detail: how each API behaves under load, how it handles part-payments and credit notes, what breaks at volume. We have not published that, and a ranking without it would be an opinion dressed as a finding.

What each of the three documents about itself is published, field by field, on Xero, MYOB and QuickBooks Online — and one difference there is large enough to mention here: Xero and QuickBooks send webhooks, MYOB does not, which changes what an integration costs to run rather than merely how it is built.

What we can do beyond that is tell you which questions decide it, because those are the same whichever package you land on.

The question that decides it: what syncs, and which way

Not "does it integrate" — every package says yes. Ask exactly what syncs, in which direction, how often, and what happens when the sync fails.

Three specifics cover most of the re-keying we are later asked to remove. Do customers sync both ways or only outbound? Do part-payments flow back to the job? Do credit notes flow at all? Anyone who actually knows the product can answer all three in a sentence.

Get those answers before you choose, because the cost of a weak accounting sync is not a one-off — it is somebody re-typing, every week, for as long as you own the software. That re-typing is the leak the whole business is built around finding.

• Customers: one way or both? One-way sync means the job system and the books drift, and someone reconciles them by hand.
• Part-payments Trades take deposits and progress payments. If those do not flow back, your job records lie about what is owed.
• Credit notes The most commonly missing piece, and the one that generates the most awkward manual work.
• Failure behaviour What happens when the sync breaks — does it queue and retry, or silently stop? Ask, and ask how you would know.
• API access tier If the API sits behind a higher plan, every future integration carries that plan as a hidden prerequisite.

What matters more than the accounting package

The pairing. An accounting package is one end of an integration; your job management system is the other, and the quality of the connection between those two specific products is what you will live with.

This is why picking the accounting software in isolation is the wrong order of operations. Choose the pair, test the join, then commit — the field test and the real-job trial on the buyer’s checklist apply just as much here.

It is also why the job management decision tends to constrain the accounting one rather than the other way round.

GST, BAS and the Australian-specific part

All three of these are Australian products or have proper Australian editions, which is not a given — plenty of otherwise good international accounting software handles GST and BAS as an afterthought bolted onto a US-shaped ledger.

That is the one place we would say something categorical: if a package cannot do BAS natively, it does not belong on an Australian tradie’s shortlist, however good the rest of it looks. Retrofitting local tax handling is expensive and it is the kind of expensive that shows up at the worst time of year — and it is one of the reasons building rather than buying so rarely pays for a compliance-shaped problem.

What we won’t claim here

No measured reliability comparison. What each vendor documents about its own limits and behaviour is now published in the capability register, and it is the register — rather than any ranking — that is actually worth reading.

We have also not surveyed Australian tradies about which package they prefer, and we are not going to write as though we had. What is on this page is the integration view from having built against all three, which is a narrower claim than most pages of this kind make.

How to actually choose, in about a week

Shortlist two accounting packages and one job management system. Ask the four sync questions above of each pairing, in writing.

Then run one real job — a messy one — through a trial of the pairing you prefer. Not a test job: a real one, with a variation on it and a part-payment. That week tells you more than any comparison page, including this one.