
Buyer's guide3 sections · 2 min readBy Dezső Mező · Published October 4, 2026.mdBuying softwareBuild vs buyTCO
The right question is not 'which is better' but 'is this process generic'. Invoicing, email, calendars, standard e-commerce — a thousand companies do them the same way, which is exactly why the shelf product is cheap and good: the cost of its development is divided by everyone who buys it. Buying standard tooling for a standard process is not compromise, it is arbitrage.
Custom earns its price where the process is the moat: the quoting logic nobody else uses, the workflow that took ten years to get right, the integration between systems that were never meant to talk. There the shelf option costs differently — not in licence fees but in workarounds, exports to Excel, and the person whose job is stitching the gap. When the workaround has a salary, custom is often the cheaper line.
01The three-question test
Would a competitor using the same tool erase your advantage? Then shelf is fine. Does the process change how you win work? Then it deserves custom. And the tiebreaker: count the workarounds — if staff maintain spreadsheets beside the tool, you are already paying for custom, just not getting it.
02The hidden bill on the shelf side
Licence fees are the visible line; the invisible ones are per-seat pricing that punishes growth, plugins that charge monthly for a checkbox, the migration when the vendor pivots, and the features you adapt your process to rather than the reverse. Total cost of the shelf is measured over years, not at checkout.
03The honest case for buying anyway
Sometimes custom is still wrong: when the process is undifferentiated, when nobody will own the result internally, or when the budget buys a bad custom job instead of a good shelf one. Custom done poorly is the worst of both worlds — the licence price plus the maintenance.
What to take away
- Standard process → shelf; differentiating process → custom.
- Count the workarounds — a spreadsheet beside the tool is already a custom-software bill.
- Shelf TCO is measured in years: seats, plugins, pivots, process-bending.
- Bad custom is worse than good shelf — budget decides which you can afford.
More from the lab
Browse all entries
What custom software actually costs
The first ninety days of a retainer, as it actually goes
Why clients get a progress page instead of status emails
Where the checkout loses the sale
Want this looked at on your own system?Start a conversation
