Building first is the most expensive way to validate. The cheaper way is to fake the product convincingly enough that real money or real commitment changes hands. Here are 5 tests that work without any code.
1. The pre-sell page (1–2 days)
Build a landing page with real pricing and a "Buy now / Reserve your spot" CTA. Charge a small deposit ($5–$20). Even one paid pre-order is gold. See Landing Page Before Product.
2. The concierge service (3–5 days)
Do the work manually for 5–10 paying users. You're the product. If they pay you to do it with no software at all, you've found demand. If they don't, no software will save it.
3. The Wizard of Oz prototype (2–3 days)
Ship a UI that looks automated. Behind the scenes, you do the work in real time. Users don't know — and crucially, don't care. Tells you if the front-end value prop lands before you build the backend.
4. The interview-to-LOI funnel (1–2 weeks)
Run 20 interviews. End each with: "If I built this in the next 8 weeks at $X/month, would you sign a non-binding letter of intent?" 5+ signed LOIs is launch-grade signal.
5. The Reddit / forum test (3 days)
Post a thoughtful question or polished mock in the niche subreddit asking if anyone else wishes this existed. Genuine engagement (not "cool idea!" but "hell yes, where do I sign up") is meaningful. Crickets is also meaningful.
What "passing" looks like
- ≥1 paid pre-order per 100 landing page visitors.
- ≥3 concierge users paying real money.
- ≥5 LOIs from cold interviews.
- ≥10 unsolicited "where do I sign up" replies in a community.
Why founders skip this
Because it's uncomfortable. It feels like fake work. It is, by design — fake work is the whole point of validation. See 48-hour validation experiments.