The question usually arrives in this form: what software should I be using to automate my business?
It is the wrong question, but understandably so, because everyone selling to trades answers it with a product. The more useful version is which of the four layers you are missing, because most trade businesses asking this already own three of them and have never had them introduced to each other.
The four layers
Every trade business, whether it knows it or not, runs on four things.
The phone layer. How enquiries get in and get sorted. For most small trades this is a mobile in a pocket and nothing else, which is the single biggest gap in the stack.
The job layer. Quotes, scheduling, dispatch, job cards, photos, sign-off. This is what people mean when they say job management software.
The money layer. Invoicing, payments, payroll, BAS. Accounting software.
The glue. The thing that moves information between the other three without you retyping it. This is the layer nobody sells you, and it is where almost all of the wasted time lives.
Read that list against your own week and the gap is usually obvious. If you are retyping the same address into two systems, your glue layer is you.
The job layer, and why you should not replace it
Australian trades are well served here. ServiceM8, Tradify, simPRO, AroFlo and Fergus all do a version of the same job, aimed at different sizes of business. simPRO and AroFlo lean towards larger operations with multiple crews and more complex scheduling. ServiceM8 and Tradify are common in smaller outfits.
Which one you are on matters far less than people think, and here is the important part: whatever you are already using is almost certainly fine.
Every one of these exposes some way to connect to it. So when someone tells you that automating your business requires moving to their preferred platform, what they are usually describing is their own convenience. Migrating job management is a genuinely disruptive project. You lose your history, your team relearns everything, and the thing that was actually annoying you, which was probably the phone or the paperwork either side of the job, is still there afterwards.
Change the job layer when the job layer is the problem. Not because you want to automate something adjacent to it.
The money layer is probably sorted
Xero and MYOB between them cover most of the Australian trade market, with QuickBooks a distant third. If you have one of them and a bookkeeper who is not complaining, this layer is done.
What accounting software will not do is run your day. It does not answer a phone, dispatch a job, chase a quote that went quiet, or remind a customer about tomorrow. People sometimes try to make it do those things and end up with a spreadsheet bolted to the side of it. That is a sign the job layer is missing, not that the accounting software is inadequate.
The glue is the layer worth paying for
Here is where the actual return is, and it is the least visible part of the stack.
Consider what happens now when a job finishes. Somebody marks it complete. Somebody else raises the invoice, retyping the details. Somebody notices a week later that it has not been paid and sends a reminder, or more likely does not. Somebody meant to ask for a review and forgot.
None of those steps is hard. All of them are manual, and each one has a queue in front of it.
Connecting them is ordinary automation. A tool like n8n, Make or Zapier sits between your job software, your accounting package and your calendar and moves things across as they happen. Job marked complete triggers a draft invoice. Invoice unpaid at seven days triggers a polite reminder. Job closed triggers a follow-up and a review request three days later.
That is the glue. It requires no AI, it runs on rules, and for most trade businesses it is worth more hours per week than anything clever. That is the general pattern, not a quirk of trades: the projects that quietly pay for themselves are almost never the ones that demo well.
The detail of the quoting and invoicing end of that chain is in a separate post on doing it without changing your accounting software, because that is the objection that stops most people.
Where AI actually belongs
Worth being specific, because the word gets attached to everything at the moment.
AI earns its place where something has to be understood rather than just moved:
The phone. A caller describing a problem in their own words, needing to be triaged and either escalated or booked. That is judgement on unstructured input, which is exactly the case for it. This is the layer most trades are missing entirely, and it is covered properly in the post on AI receptionists for tradies.
Site notes into a quote. A voice memo from the van, turned into a draft against your own price list and your standard inclusions. The system needs to have read your previous quotes to do this well, which is the difference between a novelty and something you would actually send. That distinction is the whole point of an assistant grounded in your own material.
Photos and descriptions. Turning a set of site photos into something a customer can read.
Everything else on the list above is rules. If a vendor is selling you AI to move a completed job into an invoice, you are paying a premium for something a webhook does more reliably.
The test before you buy anything
One question, and it filters most of the market.
What breaks if this vendor disappears?
If the answer is that your job history, your quote templates and your customer list are locked inside their platform, that is a real cost you should price in. If the answer is that you would need to rebuild some connections but your data is in systems you control, you are in a much better position.
The corollary is that the most valuable thing you will produce in this process is not software. It is the written description of how your business actually works: what makes a job urgent, what you include as standard, when you decline work, what your reminder schedule is. That document outlives every tool on this page, and the questions worth asking about it are the same ones you should be asking any automation partner.
The honest order
If you are starting from a mobile phone, a job management app and Xero, the order that gives back the most is:
- The phone, because missed calls are lost revenue rather than lost time.
- The glue between job completion and getting paid, because it is boring, mechanical and eats hours every week.
- Quote drafting, because same-day quotes win work that next-week quotes do not.
- Reminders and follow-ups, because they are small, cheap and nobody does them consistently by hand.
That is the same ranking as the four jobs worth automating first, viewed from the software side rather than the work side.
What it should not be is all four at once. The most expensive mistake in this whole area is buying a platform that promises to do everything, configuring twenty per cent of it, and abandoning it by March. That failure has almost nothing to do with the tools, which is a pattern worth understanding before you spend anything, and it is set out in the post on why small-business AI projects fail.
Our own rates for this work are published, in Australian dollars, so you can see what the connection layer costs before you talk to anybody.
If you want a second opinion on a stack you already have, tell me what you are running and where the retyping happens. Often the answer is that you need one connection built, not a new platform.
Frequently asked questions
What software do tradies need to automate their business?
Four layers: something that answers the phone, something that runs jobs, something that does the books, and something that connects those three. Most trade businesses already own the middle two. The gaps are almost always the phone and the connections between systems, not the job management software itself.
Do I have to replace my job management software to automate anything?
Almost never, and you should be suspicious of anyone who says otherwise. ServiceM8, Tradify, simPRO, AroFlo and the rest all expose ways to connect to them. Replacing the system your team already knows costs you months of disruption and retraining, and it rarely fixes the thing that was actually broken.
Is Xero or MYOB enough on its own?
For the books, generally yes. What accounting software does not do is run your day. It does not answer the phone, dispatch a job, chase a quote that went quiet or remind a customer about tomorrow's appointment. Those live in a job management layer, and the value comes from the two being connected rather than either being replaced.
Where does AI actually fit in a trade stack?
In the places where something has to be understood rather than just moved: a phone call that needs triaging, site notes that need turning into a draft quote, a photo that needs describing. Everything else, moving a completed job into an invoice, sending a reminder, chasing a payment, is ordinary automation and does not need AI at all.
How much should a small trade business spend on this?
Subscriptions for the job management and accounting layers are modest monthly costs you are probably already paying. The connection work is the variable. A single well-defined workflow is a small fixed-scope project. Automating an entire business at once is where budgets get away from people, which is why the sensible order is one job at a time.
Wondering what this would look like in your business? A short chat is usually enough to tell.
Let’s chat