Custom software vs off-the-shelf: when building actually pays off

Every company starts with off-the-shelf software, and usually rightly so. A subscription you can start using today beats a system built to your specification that takes three months to deliver. But that balance eventually flips. The question is not whether ready-made tools are good. It is when their convenience starts costing you more than a tailored system would.

The hidden tax of "good enough"

Research from Pendo shows that 80% of SaaS features are never used, and 53% of enterprise software licences sit effectively idle. Globally, that represents over $29 billion wasted annually on functionality nobody needs. But the larger cost is invisible: the daily friction of forcing your company's processes to fit the tool rather than the other way around. The spreadsheet bridging two systems, the manual export every Friday, the rule everyone must remember because the software cannot enforce it.

These workarounds do not appear on any invoice. But they consume time every day, generate errors whenever someone skips a step, and ensure that every new hire's first weeks begin with learning workarounds rather than doing work. Maintaining a heavily customised off-the-shelf system typically costs 22 to 25% of its value annually. For custom software, that figure drops to 15 to 20%. The difference looks modest at first. Over five years, it becomes an argument in its own right.

Five signals it's time to build

Not every friction point justifies custom software. But certain patterns keep appearing in companies that have genuinely outgrown their tools:

• You pay for several tools primarily to move data between them.
• A core workflow lives in a spreadsheet nobody is allowed to break.
• Onboarding a new person takes weeks of learning workarounds documented nowhere.
• Your competitive advantage depends on a process the tool cannot model.
• You are paying for licences to software you use perhaps 10% of.

What you actually buy when you decide to build

McKinsey analysis shows custom software delivers an average 162% return on investment over five years, compared to 74% for typical off-the-shelf enterprise implementations. Most well-planned projects reach break-even within 12 to 24 months. Projects with significant process automation often reach it faster, sometimes within 6 to 9 months.

But this is not only about finances. You buy a system that works the way your company works, not the way a vendor imagines it does. The software enforces your rules instead of relying on people to remember them. You own an asset rather than being a client dependent on someone else's roadmap. And you have the ability to change it: when the process evolves, the system evolves with it.

When off-the-shelf is still the right call

Building does not always make sense. If a process is standard in your industry, the market has a dozen mature tools for it, and your team does not perform it better than competitors, buying is almost always cheaper and faster. The danger is in assuming that because a tool covers 90% of your needs, it handles 90% of your value. In practice, that remaining 10% is often exactly where competitive advantage lives.

A useful test: if you removed every workaround and ran purely on the tool's built-in capabilities, would the company still do what clients pay you for? If the honest answer is "we would have to fundamentally change how we operate," that is precisely the moment to consider building.

A practical rule

Build where you are different. Buy where you are the same as everyone else. Your accounting is not a competitive advantage, so rent it. HR, payroll, email: standard processes get standard tools. But the workflow your clients actually pay you for is exactly where a custom system earns back its cost and keeps earning. The goal is not to build everything. It is to identify the two or three areas where software shaped to your exact process creates a durable advantage, and build deliberately there.