Founders often assume they need custom software when a tool they could buy would do the job for a fraction of the cost. Sometimes that is right; often it is not. Here is how to decide.
The short answer
Buy when an existing product solves your problem well and that problem is not your competitive advantage. Build when nothing fits, when the fit gaps are central to your business, or when the software itself is your edge. Most regret comes from building what you could have bought, or buying what you needed to build.
When to buy
- A mature product already does the job.
- The problem is common (accounting, email, CRM basics) and not where you win.
- You need it now and cannot justify a build.
- Total cost of buying (subscriptions) is clearly less than building and maintaining.
Buying is usually faster, cheaper upfront, and lower risk. Default to it unless you have a real reason not to.
When to build
- Nothing off-the-shelf fits how you actually work.
- The software is your differentiator or product.
- You need control, integration, or flexibility that tools cannot provide.
- The long-term cost or lock-in of buying outweighs building.
The "mostly fits" middle ground
The hard case is when a tool covers 80 percent. Two moves:
- Adapt your process to the tool if the gaps are minor. Cheapest path.
- Build a custom layer on top, or a custom tool, if the gaps are core to your value. (See when off-the-shelf isn't enough.)
Map each gap to how central it is. Peripheral gaps: live with them or configure around them. Core gaps: that is your case for custom.
If you decide to build
Build only the part that needs to be custom, and build it well. A product simulation is a low-risk way to start: it defines and builds a runnable foundation for the custom piece, so you commit to building only what buying cannot give you, with clarity on scope and cost.
Weighing build vs buy for a specific need? Start a project for an honest read.
Frequently asked
When should I just buy off-the-shelf?
When an existing product solves your problem well and the problem is not your competitive edge. Do not build what you can buy unless the custom version gives you a real advantage or a needed fit that nothing else offers.
What if off-the-shelf almost fits?
The 'almost fits' trap is real. Sometimes configuration or a small custom layer on top is the answer; sometimes the gaps are core and you need custom. Map the gaps to how central they are before deciding.