Test your riskiest assumptions first and cheapest — list 3-5 that would invalidate the most work if wrong

Why This Is a Rule

Every project rests on a stack of assumptions — beliefs treated as true without verification. "Customers want this feature." "The technology can scale to our needs." "We can hire the right people in time." "The regulatory environment won't change." Most projects fail not because of execution problems but because a foundational assumption was wrong, and the wrongness was discovered only after significant resources were committed.

The fail-fast principle inverts the natural project sequence. Instead of building first and discovering assumption failures later, you identify the highest-risk assumptions first and test them cheapest. The ranking criterion is invalidation potential: which assumption, if wrong, would waste the most work? That's the one you test first, because discovering its failure early costs a test; discovering it late costs all the work built on top of it.

The "3-5" constraint prevents analysis paralysis — you don't need to identify every assumption, just the load-bearing ones. The "cheapest test" constraint prevents test-design from itself becoming a major project. A one-week prototype, a customer interview, a regulatory inquiry, or a technical spike costs far less than the project it protects.

When This Fires

• Before committing significant resources (time, money, team effort) to any project
• At the start of any initiative where failure would be costly
• When expanding scope of an existing project — new scope brings new assumptions
• Complements Before committing to a one-way door, search for ways to make it two-way — trial periods, exit clauses, pilots, or phased rollouts (restructure one-way doors) by testing assumptions before committing

Common Failure Mode

Testing easy assumptions instead of risky ones: "Let's validate the technical approach first because we know how to do that." The technical approach may be the assumption you're most confident about. The highest-risk assumption might be market demand, which is harder to test but more likely to invalidate the project. Test in order of potential damage, not ease of testing.

The Protocol

(1) Before committing resources, list all assumptions the project depends on. (2) Rank by invalidation potential: "If this assumption is wrong, how much work is wasted?" The top 3-5 are your highest-risk assumptions. (3) For each, starting with the highest risk: design the cheapest possible falsification test. Not a proof — a falsification. "What's the quickest way to discover this assumption is wrong?" (4) Execute tests in order. If any assumption fails → stop and reassess before proceeding. The failed assumption may require a pivot, scope reduction, or project cancellation. (5) Only commit full resources after the top assumptions survive their tests. The testing cost is the insurance premium that protects the full investment.