Validating Startup Ideas — Confirming Demand with Customer Interviews and an MVP

#The Most Common Mistake — Validating After You Build

The trap early founders fall into most often is “let's just build it and see how people react.” They spend months finishing a product, only then put it in front of the market, and only then discover nobody wanted it. By that point, time, money, and ego are all already sunk in, and the cost of changing direction is enormous.

The point of validation isn't to confirm “I'm right” — it's to find out, as fast and as cheaply as possible, if you're wrong. There's a surprising number of questions you can answer before writing a single line of code: Who suffers from this problem? How are they getting by right now? How much are they already paying to deal with the inconvenience? All of this you can learn through conversation and observation.

#Problem Validation vs. Solution Validation — Keep the Order

Validation splits into two stages. First, problem validation: confirming that this problem genuinely exists and that people feel enough pain from it. Then, solution validation: confirming that your proposed fix actually solves that problem. The order matters — don't flip it. If the problem is fake and you build the solution first, you end up with a beautifully built product that's the right answer to a question nobody asked.

• Problem validation: Does your target customer actually experience this pain? How often, how severely? What are they using as a substitute right now?
• Solution validation: Is your approach clearly better than existing alternatives? Enough that people would drop their current habit and switch?
• You need to pass both before you can say “demand exists.” Either one alone isn't enough.

Approach problem validation through conversation (interviews, observation) and solution validation through action (landing pages, pre-orders, actual MVP usage), and your signal gets a lot cleaner. Plenty of people will say “I'd need that” — far fewer actually click, sign up, and pay. Closing that gap is the whole game of validation.

#Customer Interviews — Who, How Many, and How

Customer interviews are the cheapest, fastest, and highest-signal validation method there is. The key is “listen, don't sell.” This isn't a place to show off your idea — it's a place to dig into the other person's problem and their current behavior.

• Who: Whoever experiences this problem most often and most severely. Real target customers, not friends — people close to you tend to say nice things and muddy your signal.
• How many: There's no fixed number, but keep going until you start hearing the same pattern repeat. Usually by 5–10 people, the same complaints and excuses start overlapping.
• How: Ask about specific past behavior. Pull out facts with questions like “When was the last time you ran into this problem? What did you do about it at the time?”

#Comparing Validation Methods — Cost, Speed, Confidence

There isn't just one validation tool. Pick based on what you want to learn, and generally move from lightweight to heavyweight in stages: confirm the problem with interviews, gauge interest with a landing page or pre-orders, and confirm actual usage and willingness to pay with an MVP.

Surveys are fast and broad, but the goodwill baked into “that'd be nice to have” responses makes them weak as a standalone signal. A landing page — where you write your value proposition in one sentence and add a “sign up” button to track clicks and registrations — measures an actual action rather than words, making it a step sturdier. Attaching a small deposit or email registration to a pre-order raises the seriousness bar even further.

#Quantitative Signals — What to Measure in Numbers

The final gate of validation isn't a “feeling” — it's an action count. Goodwill is hard to measure, but clicks, sign-ups, return visits, and payments are all countable. The following signals are the raw material that turns your validation results into evidence in your business plan.

• Conversion rate: Sign-ups/registrations as a share of landing-page visits. The first-order indicator of whether your value proposition lands.
• Return visits/reuse: Does someone who tried it once come back? The core signal of genuine demand.
• Willingness to pay: Pre-order payments, conversion to paid. The strongest signal — “they actually paid.”
• Waitlist growth/referrals: Growth in your waiting list, share of signups coming from word of mouth.

#Turning Validation Results Into 'Evidence' in Your Business Plan

What reviewers and investors are looking for in a business plan isn't good writing — it's evidence. Having gone through validation means you've turned an “assumption” into a “measurement,” and that's exactly what evidence is. A business plan written without validation ends up filled with unverified assumptions like “projected revenue” and “large market opportunity,” and an experienced reviewer will pinpoint those gaps precisely.

What to check before writing your business plan is simple: can you write “evidence” right next to each claim you've made? A sentence you can't back up that way is still an unvalidated assumption. Before you submit, pick out those assumptions and, wherever possible, replace them with interview data or measured numbers — that closes off a lot of places your plan could otherwise fall apart under review.

#Frequently Asked Questions (FAQ)

Q. How many interviews do I need?

A. The standard isn't a fixed number — it's repetition. If a new person you talk to just echoes complaints and excuses you've already heard, you've found the pattern, and that usually happens within 5–10 people. If every conversation surfaces something completely different, that's a sign you haven't narrowed your target customer enough yet — tighten your targeting and keep going.

Q. How much of an MVP do I need to build?

A. Just enough to validate a single core hypothesis — the bare minimum. Not the whole product, just the one feature that tells you whether people will actually use this and pay for it. Sometimes a manual, no-code stand-in for the service (a concierge-style approach where a human handles it directly) is enough. Building more doesn't make validation more accurate — getting your answer faster makes it cheaper.

Q. Can I validate with just a survey?

A. A survey is good for broadly scanning a hypothesis, but it's weak as a standalone source of evidence. “I'd probably use that” responses rarely translate into actual behavior. The safer combination is: use a survey to set direction, an interview to dig deeper, and a landing page/pre-order/MVP to measure actual behavior.

#Wrapping Up — Validate Before You Build, Evidence Before You Submit

Good validation isn't flashy. It's the repeated work of talking to real target customers, putting something small out into the world and counting actual behavior, and dropping wrong assumptions quickly. The interviews and measured numbers you collect that way become the sturdiest material in your entire business plan.