The verdict in three sentences
In 2026, buying critical software costs EUR 15,000 to 60,000/year in licence plus configuration, while building represents EUR 50,000 to 150,000 of depreciable capex. The break-even between the two generally sits at 3-4 years, but the SaaS price hikes of 8 to 15 %/year already observed bring it closer. For a CFO, the real trade-off pits controllable capex and differentiation against flexible opex and speed.
Capex vs opex: the real comparison
| Dimension | Buy (SaaS) | Build (custom) |
|---|---|---|
| Accounting nature | Recurring opex | Depreciable capex |
| Annual cost | EUR 15,000 - 60,000 | EUR 0 licence + 15-20 % maintenance |
| Upfront cost | Config EUR 5,000-25,000 | EUR 50,000 - 150,000 |
| Time to go live | 4 - 10 weeks | 12 - 24 weeks |
| Scalability | Vendor roadmap-dependent | Full, at your pace |
| Vendor lock-in | High | None |
| Tax depreciation | No (expense) | Yes (asset) |
Build turns a recurring expense into a depreciable asset, which pleases the balance sheet; buy preserves short-term cash.
Priced decision model (5 years)
2026 order-of-magnitude projection, SaaS at EUR 35,000/year with +10 %/year, build at EUR 90,000 + 15 %/year maintenance.
| Year | Buy cumulative | Build cumulative |
|---|---|---|
| Upfront | EUR 15,000 | EUR 90,000 |
| Year 1 | EUR 50,000 | EUR 103,500 |
| Year 2 | EUR 88,500 | EUR 117,000 |
| Year 3 | EUR 130,850 | EUR 130,500 |
| Year 4 | EUR 177,435 | EUR 144,000 |
| Year 5 | EUR 228,679 | EUR 157,500 |
Break-even falls in year 3 in this scenario, then the gap widens in favour of build. The steeper the SaaS hikes, the earlier the break-even.
Risk / differentiation matrix
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
| Situation | Recommendation |
|---|---|
| Standard process + low differentiation | Buy |
| Standard process + tight budget | Buy |
| Core business + competitive edge | Build |
| High vendor lock-in risk | Build |
| Urgent need (< 8 weeks) | Buy |
| Highly specific, lasting integrations | Build |
Mini case study
Claire, CFO of an industrial SME in Amsterdam, presents the board with a planning-software trade-off. The chosen SaaS costs EUR 35,000/year with announced hikes of 10 %/year, i.e. ~EUR 228,000 over 5 years. The custom build is quoted at EUR 90,000 plus 15 % maintenance, i.e. ~EUR 157,500 over 5 years, depreciable over 5 years on the balance sheet. Net saving: ~EUR 71,000 and an owned asset. She recommends build, stressing that planning is a strategic differentiator and vendor lock-in is a 5-year risk.
FAQ
How long until build pays off? Typically 3 to 4 years in 2026, but break-even advances as soon as SaaS licence hikes exceed 10 %/year. Always model 5 years, not just year one.
Is build riskier than buying? Project risk exists (timeline, budget), but it's managed with serious scoping and milestones. Buy carries its own risk: vendor dependence, imposed hikes and a roadmap you don't control.
How do I present this to the board? Contrast depreciable capex (build) with recurring opex (buy), price the 5-year TCO, and add the strategic differentiation criterion. A break-even table convinces better than a speech.
Can we combine build and buy? Yes, often optimal: buy the commodity (payroll, accounting) and build what differentiates. This hybrid approach limits capex while protecting your edge.
How much weight for vendor lock-in? High for a core-business tool: price hikes, costly migration and roadmap dependence. If the software is strategic, build durably reduces this risk.
Let's scope your project. Send us the scope, user count and indicative budget, and we'll build your priced build-vs-buy business case for the board. Detailed quote within 48 h. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
