Build vs. buy: custom software for niche workflows

"Should we buy a platform or build our own?" is the wrong first question. The right one is: is this workflow a commodity, or is it the business? Answer that honestly and the build-vs-buy decision mostly makes itself.

When buying wins

If a thousand other companies have the same problem in the same shape, buy. Email, accounting, payroll, calendars, basic CRM — these are commodity problems, and the platforms that serve them have absorbed decades of edge cases you haven't hit yet. Building a custom version of a commodity is paying artisan prices for tap water.

Buying also wins when you need it running next week, when the team is too small to steward anything custom, or when the off-the-shelf tool covers 90 percent of the need and the last 10 percent is genuinely optional.

When building wins

• The workflow is the differentiator. If the way you process orders, onboard employees, or schedule crews is why customers pick you, bending it to fit a generic platform sands off the edge that makes you competitive.
• The audience is niche. Vendors build for the median customer. If your industry is small or your process is unusual, the median customer isn't you, and the roadmap will never bend your way.
• Per-seat pricing has become a tax. At $40 per user per month, a 25-person team pays $12,000 a year, forever, for features it half-uses. A custom tool at the same price is an asset you own instead of a subscription you rent — see our pricing FAQ for what those builds actually cost.
• The real need is glue. Often nothing on the market is missing except the connections. A thin custom layer that moves data between the systems you already own is the cheapest kind of software there is.

Count the real costs on both sides

Buying costs more than the sticker: implementation, migration, training, the workflow contortions to fit the tool, and the subscription itself, compounding annually. Building costs more than the build: hosting, updates, and someone accountable for keeping it healthy. Put five-year numbers on both columns before deciding — subscriptions look cheap in year one and expensive in year five; custom is the reverse.

Two honest cautions on the build side. Custom software needs a steward — if no one owns it after launch, it rots; that's what maintenance agreements exist for. And scope discipline matters: the winning move is to build the 20 percent of the workflow that's actually yours, not to recreate a whole platform.

The hybrid that usually wins

In practice, most of our recommendations land on a mix: keep the commodity platforms for what they're good at, and build the thin custom layer that makes them fit your business — the integration, the dashboard, the one screen that matches how your team actually works. It's the cheapest path to a stack that feels custom without rebuilding what the market already does well.

If you're not sure which side of the line your workflow falls on, that's a measurable question, not a matter of taste — a process audit answers it with numbers before any money moves.