Systems and vendors: when to build, when to buy

The question always comes back: "Should we buy a SaaS or build something custom?"

The right answer isn't in the upfront price. It's in the total cost over 3 years.

The 3-criteria rule

1. Differentiation

Is it something your customers see? That sets your company apart? If yes, consider building. If not, buy.

2. Process stability

Does the process change every week or is it stable? Custom software for volatile processes is burnt money.

3. Volume and SaaS unit cost

Many SaaS charge per user or per event. At 50 users it's fine; at 500 it's robbery.

Quick decision matrix

• Buy — common, stable, low volume: CRM, email marketing, accounting
• Build — differentiator, stable, high volume: pricing engine, unique onboarding flow
• Do nothing — common, unstable, low volume: not ready to systematize
• Hybrid — differentiator, unstable, high volume: use SaaS with API and build the custom layer on top

Hidden costs of building

• Maintenance (40% of initial cost per year)
• Bus factor (one person, one resignation, one problem)
• Opportunity (what you're NOT doing while building)

Hidden costs of buying

• Lock-in (switching SaaS in 3 years is expensive)
• Vendor integrations (monthly bill grows)
• Customization limits that show up too late

The right decision changes over time. Re-evaluate every year.