There’s a period in most small businesses where the SaaS tool you rented two years ago stops earning its keep.
It’s rarely dramatic. You don’t wake up one morning and hate the tool. It’s more that you keep finding yourself exporting to a spreadsheet, copy-pasting between two systems, or paying a premium tier for features you use twice a month because the basic tier won’t allow an API call you actually need.
That’s the drift. It costs more than most founders realise, and more than the build it was quietly avoiding.
Three signals you’ve outgrown it
1. Your team is building spreadsheets around the tool. This is the most reliable signal. If the tool had what your team needed, they wouldn’t be re-entering data in a spreadsheet to get the view they actually want. The spreadsheet is evidence.
2. The price keeps stepping up for features you don’t value. Most SaaS pricing is tiered. You need one capability from the mid-tier, so you pay for all of it. Over two years that gap compounds, and the marginal cost of each user creeps up every time you hire.
3. Integration keeps breaking. When two tools that should talk to each other don’t, and the glue is a person typing things twice, you’re paying the cost of integration anyway — just in labour rather than software. Labour is more expensive. Often the fix here isn’t a custom build at all — it’s a small automation bridging the two tools, which is a much cheaper conversation. I’ve written about the manual processes worth automating first separately.
The hidden cost of workarounds
Workarounds don’t show up on the P&L. They show up in five-minute tasks that happen 40 times a week, distributed across three people.
A team of four, each spending an hour a day working around a tool that doesn’t quite fit, burns about 1,000 hours a year between them. At modest professional-services rates, that’s the cost of a junior hire you didn’t need to make — or a custom tool that would have removed the workaround entirely.
Workarounds aren’t neutral either. They’re where errors creep in. If the person typing between the two systems is tired on a Thursday, the invoice goes out with the wrong number, and you find out three weeks later when the client queries it.
When custom is actually cheaper
Custom is often the wrong answer. I need to say that clearly, because it’s rare that a founder brings me a build question where the right move isn’t “adjust the tool you already have” or “switch to one that fits better”.
It becomes the right answer when:
- You’ve already tried the obvious SaaS options and none of them fits your actual workflow
- Your workflow is a genuine point of difference — not just a habit — and changing it would hurt you
- The monthly cost of your current tool plus workarounds is within reach of a one-off build
- You need something the market doesn’t sell because your business is unusual in a useful way
In those cases, a custom internal tool often pays back inside 12–18 months, and keeps paying afterwards because there’s no rent. When I led the automation team at CSIRO, that same pattern — replacing rented tooling and manual glue with things we owned — was worth roughly $280K a year in avoided infrastructure cost. The arithmetic is the same for a small business; the figure is just smaller.
What custom looks like in 2026
It doesn’t look like a 12-month enterprise project. That was 2015.
What it looks like now is a small, focused internal tool — usually a web app that one person or one team uses — scoped to do two or three things well. You can build one in four to eight weeks, deploy it privately, and iterate on it as the team uses it. The code is yours, the data is yours, and you’re not renting per-seat.
In practice, the small tools I build for Melbourne businesses tend to be things like:
- A client intake and matter-tracking tool that replaces a tangle of Google Sheets
- A private portal where clients submit documents and check status without emailing the office
- An internal ops dashboard that pulls from the three systems your business lives in
- A booking or quoting tool with rules the generic market products don’t handle
None of these is running a whole business. Each is one specific thing, done better than off-the-shelf can do it for your case. That’s the difference.
Starting small
The worst way to go custom is to scope the whole five-year vision, price it, panic, and not build it.
The best way is to identify the smallest version that would pay for itself in the first six months. Build that. Use it. Find out what you actually need rather than what you thought you needed. Extend it once you know.
That’s how I work with the small businesses I build tools for — smallest viable first, then grow the tool into what the business actually does, not what we predicted on day one. Starting prices are public on the services page: $5,500 for a simple marketing site, higher for internal tools depending on scope.
If you can feel the drift in your current tool — the spreadsheets, the workarounds, the creeping monthly cost — the first thing worth doing is sitting down and mapping what’s actually happening. That’s what a free audit does. If custom is the right answer, we’ll know by the end of it. If it isn’t, I’ll tell you that too.