Almost every company that comes to us with an idea for a custom system doesn’t actually need the whole thing. They need to solve one specific problem, but because “software” sounds grand, they plan to build the login, the billing, the calendar, and the chat right alongside it. That’s the most expensive way to start. Most of those pieces were solved long ago by someone else, better and cheaper than you’ll ever manage on your own. And that’s completely fine.

This lesson is about a single decision that will either make the whole project cheaper or run up the bill for years to come: what to buy off the shelf, and what to genuinely build from scratch.

Buy or build: what it all comes down to

Custom software only makes sense where an off-the-shelf solution gets in your way, not where it does the job. That sounds obvious, but in practice people get it backwards. They start with what they’d like to own, instead of starting with what can’t be bought.

Look at it this way: every feature you build yourself is a feature you then have to fix, update, and watch over yourself for years to come. A third-party off-the-shelf tool handles security, backups, and the fact that a whole team of developers is working on it. Your own code shifts all of that onto you.

So good custom software isn’t the one that does the most. It’s the one that builds only what truly has to be yours, and borrows everything else.

When to buy off the shelf: solve standard processes with a ready-made tool

A lot of things at a company look unique from the outside but are, on the inside, exactly the same as everyone else’s. Accounting, email, customer records, selling through a shop, bookings, signing contracts. Everyone does these in pretty much the same way, and there are polished tools built for exactly that.

Cheap to start. A monthly subscription is a fraction of the cost of building. You can have accounting software or a CRM running this afternoon.
Available right now. You’re not waiting weeks for development. You sign up and get to work.
Someone else maintains it. Security patches, new features, staying compliant. All of it runs on someone else’s budget, not yours.
It’s a well-trodden path. For common tools you’ll find guides, people who know them, and ready-made integrations with the rest of the world.

A rule that will save you a lot of money: don’t rebuild things that other companies have poured hundreds of person-years into. Don’t build your own payment gateway, your own calendar, or your own email system. It’s expensive, full of pitfalls, and the result will be worse than what you can buy for a few dollars a month.

When to build your own: when the process is your edge

Custom software pays off the moment that one specific process is the reason you’re better than the competition. A carpenter doesn’t need a custom email system, but a delivery company whose entire advantage is smart route planning very well might. Where the decisions that define your business happen, you don’t want to be boxed in by whatever a third-party tool happens to allow.

Build your own when at least one of these is true:

No tool fits. You tried three and had to bend each one so hard it turned into a contraption. Patching your way around someone else’s tool ends up costing more than doing it properly.
Data and integrations matter. You need systems that can’t talk to each other to actually talk, and you need your data to stay yours, not locked away at a vendor.
Per-user pricing is choking you. SaaS that costs a few dollars per person a month is cheap at five people. At two hundred people it’s a trap, and a custom solution suddenly makes financial sense.
You’re at risk of vendor lock-in. When your whole business rests on one third-party tool, a price hike or a shutdown is your existential risk, not theirs.

The hybrid path: you almost never build everything

Here’s the secret most people won’t tell you. Real systems are neither purely bought nor entirely custom. They’re a thin custom layer that glues ready-made building blocks together.

You handle payments through Stripe. Login through an existing service. Email through an existing tool. Data in an ordinary database. And you build from scratch only the one thing that is your heart: the logic, the planner, the calculator, or the workflow that the whole thing exists for.

A good brief doesn’t say “build us a system.” It says “we’ll buy this off the shelf, plug that in, and build from scratch only the one thing nobody else has.”

This approach gets you software faster, cheaper, and more securely. You build ten percent and borrow the other ninety from people who do it better.

Questions to ask yourself before you decide

Before you send out a request for a “custom system,” go through this. By all means, point by point, on paper:

Is this specific thing what sets us apart? If it isn’t the reason we’re better, buy off the shelf. It’s worth building a competitive advantage, not overhead.
Does a good tool already exist? Honestly try three. Not the first one that shows up on Google. You might find the problem is already solved for a subscription.
What would it cost us to leave that SaaS one day? Can we get our data out in a usable form? This one question decides whether an off-the-shelf tool is a convenience or a future trap.
Can we carry the running cost of what we build? Custom software isn’t a finished thing, it’s a commitment for years. Who maintains it once we’re busy with other things?

If those answers tell you that what sets you apart can’t be bought, you have the green light for a custom build. If they tell you that you just want to own it, save your money and buy the ready-made box. In the next lessons we’ll break down what such software is actually made of, and how to get from an idea to a first usable version.