For a long time, building job management software was expensive. You needed a team, servers, years of learning, and enough money to survive the years before anyone paid you. That was the moat. Not the idea. The cost and the time.
That moat has mostly gone. A competent developer with AI tools, a few open-source libraries and a cloud account can put a working job management demo in front of you in a weekend. A tradie with a clear idea and a subscription to a no-code platform can do something similar in a month.
That sounds like good news. More competition, lower prices, more choice. It is not good news by itself. The part that got cheap is the part you can see. The expensive part is still expensive, and if nobody paid for it, you will.
You can already see the result in the marketing. "Built by a tradie, for tradies" is offered as a credential. It is closer to a warning. The sentence you should want to hear is "built by experts, refined by tradies."
What got cheaper, and what did not
The visible layer of job management software is now cheap to produce. Login screens, client lists, job cards, drag-and-drop scheduling, a dashboard, a mobile view. AI assistants can generate a lot of that quickly, and no-code tools can wire it to a database without a computer science degree.
The invisible layer did not get cheaper:
- deciding what happens when two staff members edit the same job at the same time;
- keeping quotes, job costs and invoices consistent when a variation is approved;
- getting permissions right so a subcontractor sees the job but not the customer's payment history;
- migrating five years of records without duplicating clients or losing job history;
- handling GST, progress claims and invoice numbering correctly;
- restoring a backup after someone deletes the wrong thing;
- answering the phone at 7am when a job has disappeared from the schedule.
None of that demos well. All of it is most of the product.
AI raises the floor. It does not raise the ceiling. It helps you build a working screen, not a system that survives a real business on a bad Tuesday.
The moat was never really the code
The old moat was money and time. The real moat was always judgement: knowing which edge cases matter, and having the discipline to handle them before a customer finds them.
A useful job management system encodes hundreds of small decisions. What happens to the schedule when a job runs three hours over? Does a quote revision keep the old version? Can a technician edit a price after the customer has approved it? Which records belong to the client, and which belong to the site? What does "complete" mean, and who is allowed to say it?
Those decisions are not technical trivia. They are the business. A software team that has never run a trade business will often make them badly because it has never seen the consequences. A tradesperson who has never run a software product will make them narrowly because they have only seen their own business.
When the cost of a demo falls to near zero, the market fills with products that look finished because the screen is finished. The buyer cannot see the missing decisions until they hit one.
You become the QA department
This is the part that should worry you. When software is cheap to start but expensive to finish, the finishing gets done by the first real customers. That is you.
Some examples of what that looks like in practice:
- A quote revision overwrites the approved version because history was never modelled. You send the wrong price.
- A job is reassigned, but the calendar and the SMS reminder still point at the first technician. Nobody turns up.
- Duplicate client records split the history, so the office quotes a repeat customer as new.
- A permissions bug lets a subcontractor see every customer's contact details.
- Photos upload in the office but fail silently in a dead spot, so the evidence for a variation is gone.
- The export button produces a spreadsheet that drops job notes, and you find out when you try to leave.
A bug in a game is annoying. A bug in job software is an unbilled job, a missed appointment, or a privacy breach.
None of this means the people building these tools are dishonest. Most are not. They are learning in public, on live businesses, because the tools now let them. The problem is that the learning is funded by your jobs. In practical terms, you are not buying expertise. You are funding someone's education, and the tuition is your admin.
Cheap to build also means cheap to abandon
There is a second cost that buyers rarely think about. When a product is cheap to start, it is also cheap to walk away from.
A vendor with a weekend prototype and a subscription page has very little sunk cost. If growth stalls, the founder goes back to their day job, the domain lapses, and your records sit in a system nobody is maintaining. Even if the software keeps running, nobody is fixing the bug that lost you a job last month.
Longevity is not a guarantee of quality. Plenty of old software is bad. But a vendor that has been supporting the same product for years has demonstrated something a marketing page cannot: it has survived real customers, real incidents and real payroll cycles.
The car analogy
I love working on cars. I have done my own brakes, chased electrical faults, and spent weekends on jobs that should have taken an hour. I would still not put a sign on the fence and charge strangers to fix theirs.
Not because I cannot do the job. Because doing it once in my driveway, on my car, at my pace, is a different activity from doing it every day for customers who need the car back by 5pm. The second one needs insurance, a warranty, a calibrated torque wrench, and someone who will answer the phone when it goes wrong. Enthusiasm is not a warranty.
I can also tell a mechanic exactly what is wrong with my car: which brake is grinding, when the shudder started, and what it sounds like on a cold morning. That makes me a useful test driver. It does not make me the mechanic.
Software is the same. Building something that works for your own business is a real achievement. It is not the same as software another business can trust with its quotes, invoices and customer records. The gap between "it works on my jobs" and "it works on yours" is where a lot of small software companies quietly die, usually after a customer has already moved their business onto it.
"Built by a tradie" is a warning, not a feature
"Built by a tradie" is probably the worst thing you can hear about a job management system. It tells you who had the idea. It does not tell you who is responsible for the data model, permissions, backups, security or support. In the worst cases, it is used to skip that question entirely. The founder's trade experience becomes the answer to every technical concern, and the customer pays for the gap.
The sentence you want is "built by experts, refined by tradies." Experts build the product. Tradies use it hard, find what is wrong, and tell the builders what the work actually needs. That loop produces good software. Remove either half and you get a different failure. Without the experts, you get a demo that cannot survive real use. Without the tradies, you get polished software for a business that does not exist.
I am not arguing that tradies should stay out of software. The opposite. Trade experience is what makes the refinement valuable. But it is an input to the product, not a substitute for engineering.
There is a third group now: people with neither deep trade experience nor deep software experience, using AI to close both gaps at once. That is where you get a product that looks finished before it is. Some will grow into real companies. Most will not, and the customers who signed up early will carry that cost.
Questions that separate a product from a prototype
You do not need to become a software expert. You need to ask for evidence that someone else is one.
- Who writes the software, and is that their full-time job?
- What happens when two people edit the same job at once?
- Can I export everything, including notes, photos and history, in a usable format?
- How are permissions handled for office staff, technicians and subcontractors?
- Who supports the product, and what happens after hours?
- When was the last backup restore actually tested?
- How do you migrate data from my current system without duplicating records?
- Where is my data hosted, and who can access it?
- What happens to my data if I leave, or if you close the company?
- Can I see the system handling real jobs end to end, or speak with a business already using it?
A serious vendor will have answers. They may not be perfect answers, but they will be specific, and they will include limitations. A prototype will answer with a feature list.
The red flags are consistent:
- no real export path;
- no clear support process or escalation path when something breaks;
- "we can build anything you need" with no product boundary;
- no contract, no data processing terms, no clear ownership of your data;
- no customers outside the founder's own business;
- pricing that cannot possibly fund support and development.
What I would actually pay for
This is not a neutral question for me, so I will be direct about the bias. I work at Proanalytica, and ToolTorq is our product.
ServiceM8 is the obvious established example of the right split. It grew out of a locksmith business Kim Ford bought in 2009, so the domain knowledge is real. But ServiceM8 is a software company. It shipped the first version in 2010 and has been building and supporting the product ever since. The locksmith business supplied the problem. The software team built the system.
Fergus is similar. Dan Pollard founded it in New Zealand in 2011 after years in the trades, and the company has spent more than a decade engineering and supporting the product. It now supports customers across New Zealand, Australia and the UK. The trade experience shaped the product. It did not replace the engineering.
ToolTorq is our version of the same split. It started inside client work rather than on a whiteboard: trade businesses kept hitting the same office gaps around quoting, client history, scheduling, staff leave, documents, job margin and communication, and we kept building the same workflows by hand. Turning that work into a dedicated CRM and job management platform let us invest properly in the underlying data model, permissions, backups and product boundaries instead of treating every client as a one-off build.
The reason I trust it is not that we are clever. It is that we engineer software for a living, and our team works inside Australian trade and field-service businesses every week. Every decision in ToolTorq is tested against real office admin, real field staff and real client systems rather than a weekend pitch.
That is the split in practice: built by experts, refined by tradies. If you want an established platform with a decade-plus track record, ServiceM8 and Fergus both have real engineering teams and long production histories. If you want a modern CRM and job workspace shaped directly around how Australian trade offices work, that is what ToolTorq is built for, and we will tell you honestly if it is not the right fit. If ServiceM8 alone already does the job for your crew, stay there—you do not need extra software just for the sake of it.
The cheap moat is your problem now
Cheaper software is good for choice. It is not automatically good for buyers. When the barrier to entry drops, the average quality of a new product drops with it, at least for a while.
That does not mean you should cling to a spreadsheet. It means the demo is no longer the test. The test is what happens after the demo: the edge cases, the support, the export, the security, and the willingness of the vendor to tell you what their product does not do.
The phrase to listen for is "built by experts, refined by tradies." If you only hear the second half, ask the questions above. If a new tool cannot answer them, wait. Your jobs are not a development environment.
Proanalytica works with Australian trade and field-service businesses on ServiceM8 setup, custom workflows and ToolTorq. If you are comparing job management software and want a second opinion, talk to us.
Related reading
- Job management software for tradies: when to upgrade
- ServiceM8 customisation: when tradies need more
- ToolTorq: the CRM layer Proanalytica is building
