It's the question I get more than any other: "Should we just buy a tool for this, or have something built?" And the honest answer surprises people — for most processes, you should buy. Off-the-shelf software is cheaper, faster and someone else keeps it running. A developer who tells you to build everything custom is usually selling hours.
But "buy by default" has a failure mode of its own: you end up with five subscriptions that each solve 80% of a problem, and a person whose real job is copying data between them. So the question isn't buy or build in the abstract — it's knowing which one your specific situation calls for. Here's the framework I use.
When buying off-the-shelf wins
For generic, well-understood processes, standard SaaS is almost always the right choice. Accounting, CRM, payroll, email marketing, project management — thousands of companies need exactly the same thing, so a vendor can spread the development cost across all of them and charge you a fraction of what building it yourself would cost.
Buy when:
- The process is a commodity. If your invoicing works like everyone else's invoicing, you gain nothing by building it from scratch.
- You need it now. A SaaS tool is live this afternoon. Custom software takes weeks to months.
- The vendor handles maintenance. Security patches, uptime, new features — that's their problem, not yours. With custom software, maintenance is a real recurring cost you carry.
- The total cost stays low. A few users on a €30–€80/month tool will almost never justify a custom build.
If you're early in figuring out what software your business even needs, start by reading where to start with digitalising an SMB — the answer there is also "buy first, build only where you have to."
When building custom wins
Custom software earns its keep in a narrower set of cases than vendors would like you to believe — but when it fits, the payoff is large.
- The process is your edge, not a commodity. If the way you do something is part of why customers choose you, forcing it into a generic tool throws away your advantage. Standard software makes you work like everyone else.
- No tool fits, so you're paying in glue. Three systems that don't talk to each other, an employee manually moving data between them — that's a hidden cost you already pay, in salary and errors. Sometimes the fix is an integration between the tools you have; sometimes it's a custom application that replaces the lot.
- The subscription math flips. SaaS is cheap for a few users and expensive at scale. €2,500/month across forty seats is €30,000 a year — every year, with prices that only go up. A custom tool is a one-time build plus modest upkeep.
- You've outgrown the workaround. The classic version is a business run on spreadsheets that have quietly become mission-critical. I wrote a whole piece on when the move from Excel to custom software is worth it.
The numbers, side by side
Take a 30-person company that needs a system to manage a specific operational workflow.
- SaaS option: €40/user/month × 30 = €1,200/month, or €14,400/year. Live in days. But it covers maybe 70% of what you need, so two people spend a few hours a week working around the gaps.
- Custom option: €35,000 to build, plus roughly 15–20% per year in maintenance (~€6,000). Higher upfront, nothing per seat, and it fits the workflow exactly.
Over five years the SaaS path is ~€72,000 plus the cost of the manual workarounds, and you own nothing. The custom path is ~€59,000 and you own an asset built around how your business actually runs. The break-even is usually somewhere between year two and three — sooner if the SaaS gaps are forcing real manual labour. For more detail on what drives the build figure, see what custom software costs.
The middle ground people miss
Build vs. buy isn't binary. The most cost-effective answer is often a mix: buy the commodity parts, build the bit that's actually yours, and connect them.
- Buy + integrate. Keep your accounting and CRM SaaS, and build a thin integration layer so they finally talk to each other. Far cheaper than replacing anything.
- Buy + automate. Keep the tools, but wrap a custom automation around the manual steps between them.
- Build only the core. Build the one workflow that's your competitive edge, and buy everything around it.
The questions that actually settle it
When a client can't decide, these four questions almost always break the tie:
- Is this process generic or specific to us? Generic leans buy. Specific leans build.
- What does the gap cost us today? Add up the manual hours and errors caused by tools that don't quite fit. That number is the real budget for a custom fix.
- How will the cost scale? Per-seat SaaS that looks cheap at ten users can be painful at a hundred.
- What happens if the vendor changes course? Price hikes, removed features, a shutdown — buying means you don't control the roadmap. Weigh that against the upkeep a custom build asks of you.
One warning in the other direction: don't build to avoid vendor dependence if the thing you're replacing is a commodity. Rebuilding accounting software to "own it" is how good budgets get wasted.
Not sure which side you're on?
Most companies sit somewhere in the middle and don't need nearly as much custom software as they fear. I'll happily look at your situation and tell you honestly where buying is smarter — even when that means I build nothing. Describe the process that's frustrating you, and let's work out whether it's a buy, a build, or a bit of both.