Custom software or off-the-shelf SaaS, and how to decide

A growing company reaches a point where a subscription tool no longer fits how it works, and someone asks whether it is time to build its own. The honest answer depends on a few things that can be checked, not on which option looks cheaper in a spreadsheet drawn before either exists.

The short answer#

Use ready-made software for work that is the same in every company: payroll, bookkeeping, email marketing, booking. Build custom software where how you do something is part of why customers choose you, or where the ready-made tool forces a workaround that costs more than the tool saves. Most companies end up with both, joined by integrations.

Three checks settle most cases. How many hours a week do people spend working around the current tool? Can the product close the gap with a setting, a plugin or its API? Could you leave the vendor with all your data if the price doubled? If the answers are "many", "no" and "not easily", start thinking about custom software. If they are "few", "yes" and "yes", stay with the product.

When off-the-shelf is the right choice#

• The process is standard: payroll, general ledger accounting, email marketing, calendar booking. Doing it differently earns you nothing.
• You need it working soon. Configuring a product is faster than designing one.
• The team is small and the requirements are plain.
• A mature product already covers what you need, and the remaining gap can be closed with a setting, a plugin or an integration.

When custom starts to make sense#

• The process is how you compete, such as your own pricing rules, routing or business-customer terms, and the tool makes you change it.
• People work around the tool: exports to spreadsheets, the same data typed into two systems, rules kept in someone's head. Count the hours per week. That is the most honest cost signal you will get.
• Per-user or per-record pricing has grown with the company while you use a small part of the features.
• Several systems must work together and the vendor's integration is limited or unreliable.
• A customer contract or sector rule sets where data lives or who can see it. Check the actual requirement first, because a vendor may already meet it.

What each option really costs#

The amounts depend on scope, so none are given here. What drives them is more useful.

Ready-made software: a subscription that usually grows with users or volume, the work of configuring it and training people, integrations with your other systems, price changes at renewal, and the cost of leaving, which means exporting your data and rebuilding the process elsewhere.

Custom software: design and build, hosting, security updates and maintenance after launch, documentation, and a person or partner who understands the code. The cost does not end at launch. Software without an owner decays, so plan for its upkeep from the start.

Questions that settle it#

1. Which steps do people do by hand around the current tool, and how many hours a week does that take?
2. Can the ready-made product cover the missing part through settings, plugins or its API?
3. If the vendor doubled the price or dropped a feature, what would you do? Can you export all your data in a usable form?
4. Who would own and maintain a custom system in its third year?
5. Is there a middle path: keep the product for the standard part and build only the missing piece beside it, connected through its API?

What this article does not cover#

It does not compare specific products and gives no cost figures. For regulated data, confirm the requirements with a specialist before choosing either route.

How Nevo approaches it#

Nevo builds custom systems and integrations around software companies already use. If you are weighing this decision, see software development.