When Custom Software Beats Off-the-Shelf

The short version

Most businesses should buy off-the-shelf software, and I'll say so when asked. Custom becomes the better answer when you're paying people to work around a tool that doesn't fit — re-typing data between systems, maintaining a spreadsheet beside the app, or paying per seat for features nobody uses. The tell isn't the software budget, it's the hours your staff spend compensating for it. Building small tools got substantially cheaper in the last few years, which moved that line.

I build custom software for small businesses, so treat what follows accordingly. But the honest starting position is this: most of the time you should just buy the thing.

When off-the-shelf is the right answer

If your process is genuinely the same as everyone else's, someone has already built better software for it than I would, and they've had a hundred customers find the bugs first.

Accounting and payroll are the clearest examples. Tax rules change every year, compliance is unforgiving, and there's no version of your business where invoicing math is a competitive advantage. Buy QuickBooks. Same goes for email, calendars, and file storage.

The general test: if a product does 90% of what you need and costs less than a few hundred a month, the remaining 10% is almost never worth building around.

The tell: you're paying people to work around your software

The situation flips when the workarounds start costing more than the software.

You know the ones. Somebody exports a report and re-types it into a second system. There's a spreadsheet living beside the app because the app can't track the one thing you actually care about. New hires get a verbal explanation of which fields to ignore and what the naming convention means. A job gets entered three times because quoting, scheduling, and invoicing don't talk to each other.

None of that shows up on the invoice. It shows up as hours, mistakes, and the specific irritation of watching a competent person do something a computer should be doing. That's the real cost of software that doesn't fit, and it's the number worth comparing against, not the subscription.

What you actually get from building it

It fits the process you already have. Off-the-shelf software makes you adapt your business to the vendor's idea of how your trade works. Custom goes the other direction. If your quotes need a field for site access notes because half your jobs are in buildings with no parking, you get that field — you don't get a "notes" box you agree to use for that purpose.

You own it. No per-seat pricing that grows every time you hire. No annual increase you have no leverage over. No vendor getting acquired and sunsetting the product with twelve months' notice. Your data sits in your database, in a form you can read. It's the same argument as running your own office server — the things your business depends on should answer to you.

Support is a person who knows your business. When something breaks you're not describing your setup from scratch to a tier-one agent working from a script. You message the person who wrote the code, and they already know why that field exists.

Changes happen on your timeline, not a roadmap. With a SaaS product, a feature request goes into a queue weighted by how many other customers want it — which for a specific request from a small customer means never. With something built for you, "can it do X" is usually a conversation and then a few days, not a ticket that dies quietly.

It grows with you instead of against you. Off-the-shelf tools tend to fail at the edges: you outgrow the cheap tier, or the enterprise tier assumes a company ten times your size. Software built around your actual operation gets extended when the operation changes.

The comparison, briefly

Off-the-shelf Custom
Fit You adapt It adapts
Cost Per seat, grows Build, then flat
Support Ticket queue Whoever built it
Features Vendor's roadmap Yours
Growth New pricing tier Just extended
Data Their system Your database
Best when Process is standard Process is the business

To be clear about the monthly part: whether there is one depends on the shape you pick, and that's a decision you get to make.

A hosted web app is a monthly subscription — I run it, and small improvements are part of it, with larger pieces of work quoted by size. A desktop app running against a database on your own machine is a one-time build cost with no monthly at all: bug fixes are free for as long as you use it, new features are quoted when you want them, and backups become your responsibility. Cloud backup is available as an add-on to either, priced on storage. And if you'd rather the whole thing lived in your building, a server in your own office can run it.

The distinction that actually matters against off-the-shelf isn't monthly versus not. It's that none of these are priced per seat. Adding your tenth employee changes the bill on a subscription and doesn't change it on any of these.

Why this became worth considering

Custom software used to be the expensive option by a wide margin, which is why it was reserved for companies large enough to have an IT department.

That gap narrowed considerably. Better tooling — including AI-assisted development, used carefully — has cut the time to build a small, well-scoped tool enough that projects which would have been a month of work are now a fraction of that. It hasn't made large systems cheap, and anyone telling you otherwise is selling something. But it moved the line far enough that a five-person business can now reasonably justify a tool built for it, which was not true a few years ago.

SaBooks is my own version of this. I was juggling three tools to send a proposal and get paid for one job, so I built the small thing that did exactly that and nothing else.

The honest downsides

You're depending on a smaller number of people. That's a real risk and the answer is boring: the code and documentation are yours, written so another developer can pick it up. Anyone who won't hand that over is selling you a hostage situation, not software.

You also don't get a 24/7 support desk, the build cost lands upfront instead of spread thinly across monthly payments, and somebody has to maintain it once it exists. Those are real, and they're why the answer is genuinely "buy it" more often than "build it."

Is it worth it for you?

The quickest way to find out is to count the hours. If someone in your business spends a few hours a week doing something a computer should do, that's the number to weigh — and it's usually larger than people expect once they add it up.

If you're trying to work out what any of this should cost, how IT work is actually priced covers the models you'll be quoted under and what to ask before you sign anything. And if you're a founder rather than an established business, the startup version of this question is a different one.

If you've got a process held together by a spreadsheet and a good memory, get in touch. I'll tell you plainly whether it's worth building, including when the answer is that you should keep paying for the thing you've already got.

← Back to Insights