At a startup, you don't have a 20-person QA team. You have 2 engineers and a deadline. You can't test everything.
The question isn't "should we test?" — it's "what do we test first?"
The Risk-Based Testing Pyramid
Forget the traditional testing pyramid (unit > integration > E2E). For startups, I use a risk-based approach:
Priority 1: Test things that lose money. Payment flows, subscription management, billing calculations. A bug here costs real dollars and real customers.
Priority 2: Test things that lose data. Database migrations, data exports, backup/restore. A bug here is catastrophic and often irreversible.
Priority 3: Test things that lose trust. Authentication, authorization, password reset, email delivery. A bug here makes users question your security.
Priority 4: Test everything else. UI interactions, edge cases, performance, accessibility. Important but not existential.
The Minimum Viable Test Suite
For a typical SaaS startup, here's what I'd set up in week 1:
\



