Every growing business eventually hits the same wall: five, six, seven SaaS subscriptions, each solving one slice of the problem, none of them talking to each other properly. The instinct is to add an eighth tool. We usually push back on that instinct.
The stacking problem
SaaS tools are built for the average customer, not for you. That's fine for accounting, email, and other genuinely generic problems — buy the best tool and move on. It stops being fine when your actual process (how your warehouse tracks batches, how your sales team qualifies leads, how your production floor reports defects) has to bend around a tool that was never designed for it.
The result is familiar: a CRM that half the team avoids because it doesn't match how they actually sell, a workaround spreadsheet living next to an expensive platform, an integration that breaks every time one vendor ships an update you didn't ask for.
When a small custom app wins
We build a custom app when three things are true at once:
- The workflow is specific enough that no off-the-shelf tool fits it without heavy compromise
- The workflow is used often enough that the compromise costs real time every week
- The scope is small enough to ship a useful first version in weeks, not months
That last point matters most. A custom app doesn't have to mean a year-long enterprise project. Built with the right stack — Supabase for data, a lightweight frontend, n8n for the glue to everything else — a focused internal tool can ship fast and grow only as far as it's actually needed.
What you get instead
Software that matches your process instead of the other way around. One source of truth instead of a spreadsheet fighting a platform. And no monthly fee scaling with seats you don't need, for the parts of your operation that are actually yours to own.
The point isn't "SaaS bad." Most of what a business needs should stay bought, not built. But the exceptions — the workflows that are genuinely yours — are usually worth a small, well-scoped app rather than another subscription that almost fits.