How do I validate a startup idea without building anything?
Sell before you build. The goal is to prove that someone will commit money, time, or attention to the outcome you promise — without production software. Use customer interviews to confirm the pain, a landing page to measure interest, a manual or concierge version to deliver the value by hand, and pre-sales or deposits to confirm willingness to pay.
If no one commits at the manual stage, code will not fix that. The strongest validation signal is not "that sounds interesting" — it is someone handing you money or blocking time on their calendar before the product exists.
Why this matters
Most failed startups build something nobody asked for. The code was fine — the demand was not there. Validation without building forces you to confront demand before investing months of development. If you can sell the outcome by hand, you know exactly what to automate. If you cannot sell it by hand, no amount of engineering will help.
Five ways to validate without code
- Customer interviews (10–20 conversations): Ask how they solve the problem today, what it costs them, and what they have tried. Do not pitch your idea — listen for patterns. If 3+ people describe the same pain and workaround, you have a signal.
- Landing page test: Build a single page describing the outcome (not the features). Add a "Join waitlist" or "Pre-order" button. Spend $100–$300 on targeted ads. If strangers click, the positioning works. If they don't, rewrite the value proposition.
- Concierge MVP: Deliver the value entirely by hand. If the product would automate reporting, do the report manually for 5 clients. If they pay and come back, you have validated the outcome. You also learn every edge case before writing a line of code.
- Pre-sales or deposits: Ask potential customers to pay upfront for early access. Even $50 deposits are a stronger signal than 1,000 email signups. Money committed is demand confirmed.
- Competitor teardown: Find the closest existing solution and read their 1-star reviews. The complaints tell you what to build differently. If the competitor has paying customers despite bad reviews, the demand is real — the execution is the gap.
What counts as validated
- 3+ people pay for a manual or early version before software exists
- A buyer describes the problem unprompted with a specific cost or frequency
- A landing page converts strangers at 5%+ signup rate from cold traffic
- Someone signs a letter of intent, pilot agreement, or deposits money
What does not count as validation
- "My friends think it's a great idea"
- Upvotes on a Reddit post
- Email signups with no commitment attached
- A survey where people say they "would" use it
Related questions
Get ranked business directions with validation steps tailored to your background and market.
Open the workspace →