SaaS won the default for most business software. The question is no longer "should we consider SaaS?" but "at what point does SaaS stop being the right answer?" Here's a clear decision framework.
Three conditions under which custom software consistently beats SaaS
1. Your process is genuinely unique
SaaS tools are built for the median user in a market segment. They make assumptions about how your industry works, what your workflow looks like, and what data you care about. When your actual process deviates significantly from those assumptions, you spend a disproportionate amount of time working around the tool rather than with it.
Signs this is happening: you've built complex spreadsheet systems to compensate for what the SaaS can't do; your team spends 20%+ of their time on manual data transfer between tools; you regularly ask your SaaS vendor for features and are told they're on the roadmap; you've hired a dedicated person to manage your tool stack.
2. The SaaS is a competitive asset you can't afford to share
If the workflow you're automating is the thing that makes you better than your competitors, you probably shouldn't run it on the same SaaS tool as your competitors. A logistics company whose routing optimisation is its main competitive advantage should own that system, not license it from a vendor who sells the same tool to every logistics company in the market.
3. Per-unit economics flip above a certain volume
Many SaaS tools charge per seat, per transaction, or per API call. At low volume, this is efficient — you only pay for what you use. At high volume, you're paying a percentage of your revenue or cost to a middleware vendor. The math often changes at 50–200 seats or 100,000+ monthly transactions.
The hidden costs of SaaS at scale
SaaS pricing is transparent at small scale. At scale, the total cost often surprises operators:
- Seat expansion: Tiered pricing means your cost per seat often increases as you grow — the opposite of typical economies of scale.
- Feature add-ons: The feature you need for your use case is usually in the next pricing tier. Enterprise features (SSO, audit logs, advanced permissions) are almost always priced at a significant premium.
- Data portability: Extracting your data from a SaaS tool for analytics, migration, or audit purposes often requires the premium API tier or a custom extraction project.
- Integration costs: Building and maintaining reliable integrations between SaaS tools has real engineering costs, even with iPaaS tools like Zapier or Make.
A framework for the decision
Answer these five questions:
- Is a good SaaS tool available for this specific use case? (If no, the decision is made.)
- Is your process close enough to the tool's assumptions to work without heavy customisation?
- What is the 24-month TCO of SaaS vs custom at your expected volume?
- Does this workflow represent a competitive differentiator you want to own?
- Do you have the engineering resource (internal or outsourced) to build and maintain custom software reliably?
If you answered yes to questions 1 and 2, and no to 4, use SaaS. If you answered no to 2, yes to 4, and yes to 5, build custom. Everything else is a judgment call.
For a detailed cost comparison for your specific project, use our Build vs Buy Decision Tool and Cost Estimator. See also: Custom Software vs Off-the-Shelf.