Stop Validating Your Startup Idea. Start Selling It.

"Have you validated it?"

If you've ever posted a startup idea anywhere on the internet, you've gotten this response. It's the default advice. The safe advice. The advice that sounds smart and says nothing.

I've watched hundreds of indie hackers go through "validation" and here's what I've noticed: the ones who spend months validating rarely ship anything. The ones who just build and sell? They're the ones making money.

Let me be clear about what I mean: I'm not saying research is useless. I'm saying that what most people call "validation" is just procrastination wearing a business costume.

What validation actually tells you

Let's be honest about what common validation methods actually prove:

Landing page with email signups. This proves people will give you an email address. It does NOT prove they'll pay. I've seen products with 3,000 email signups convert exactly 8 paying customers. Email is free to give. Money is not.

"Would you pay for this?" surveys. This proves people are polite. In every survey ever, the majority says "yes, I'd pay for that!" In reality, less than 5% of those people ever open their wallets. What people say and what people do are completely different things.

10 user interviews. This proves the problem exists. It does NOT prove your solution is the right one. You can correctly identify a real problem and still build the wrong product, at the wrong price, for the wrong segment.

Competitor analysis. This proves a market exists. It does NOT prove there's room for you. Markets can be large and still be winner-take-all.

None of these are useless. But none of them are the definitive signal the validation crowd pretends they are.

The only validation that matters

There is exactly one form of validation that actually works: someone gives you money.

Not a promise to pay. Not "I'd definitely buy this." Not an email signup. A credit card transaction. Dollars in your Stripe account.

Everything else is a proxy. Some proxies are better than others, but none of them are the real thing.

This isn't a radical idea. Daniel Vassallo has been saying this for years. His whole "small bets" philosophy boils down to: build something small, charge money, see if anyone pays. If they do, keep going. If they don't, try the next thing.

The 2-week sprint

Here's the framework I recommend instead of traditional validation:

Week 1: Build the smallest possible version.

Not a landing page. An actual product. It can be ugly. It can be missing 80% of the features you eventually want. But it needs to DO the core thing and have a way to accept payment.

If you can't build a working version in a week, your scope is too big. Cut features until it fits.

Week 2: Sell it to 20 people.

Not "launch" it. Sell it. Directly. One person at a time.

Go to the communities where your target customers hang out. DM people who have the problem your product solves. Show them the product. Ask them to pay.

This is uncomfortable. That's the point. If you can't sell something face-to-face (or DM-to-DM), a landing page won't sell it for you either.

After 2 weeks, evaluate:

• If 3+ people paid: you've got something. Keep going.
• If 0 people paid despite genuine effort: either the problem isn't painful enough, your solution doesn't fit, or your audience is wrong. Adjust or move on.
• If people were interested but didn't pay: your pricing might be wrong, or your product might need one more key feature. Talk to the interested ones and figure out what's missing.

Why this works better than validation

Speed. The entire cycle takes 2 weeks, not 3 months. If the idea doesn't work, you've lost 2 weeks instead of a quarter.

Real signal. Money is the strongest signal in business. Everything else is noise by comparison.

You build a product. Even if the idea fails, you now have code, experience, and maybe even a few customers you can talk to about what they DO need.

You build sales skills. The #1 skill for a solo founder isn't coding. It's selling. Every failed sales conversation teaches you something about your market, your positioning, and your communication.

But what about building the wrong thing?

This is the main argument for validation: "What if you build something nobody wants?"

My response: what if you validate something for 3 months and STILL build the wrong thing? Because that happens all the time. Surveys and interviews give you hypothetical data about hypothetical behavior. The product you build based on that data often misses the mark anyway, because the gap between what people say and what they do is enormous.

At least when you build something in 1-2 weeks, the cost of being wrong is minimal. You spent 2 weeks, learned that the market doesn't want this specific solution, and you can try again with real feedback.

Compare that to: you spent 3 months "validating," felt confident, spent 2 months building, launched, and... nobody bought it. That's 5 months wasted instead of 2 weeks.

When to actually validate

I'm not saying never do research. There are specific situations where upfront validation makes sense:

• When the build cost is high. If your product requires hardware, a team, or $50K+ in development, yes, validate first. The 2-week sprint is for software you can build yourself.
• When the market is completely unknown to you. If you've never interacted with your target customer, spend a week talking to them before building. But a WEEK, not a quarter.
• When regulations are involved. If your product touches healthcare, finance, or legal compliance, you need to understand the regulatory landscape before building.

For everything else? Build it. Sell it. Learn from it.

The real validation trap

The deep reason people over-validate isn't rational. It's emotional. Validation feels productive without the risk of rejection. You can spend months "researching" your idea and never face the moment where someone says "no, I won't pay for this."

But that moment of rejection is the most valuable data point in entrepreneurship. Every "no" teaches you something. Every "yes" proves something. And you can't get either one from a survey.

Stop validating. Start selling. The market will tell you everything you need to know - faster, more honestly, and more cheaply than any validation process ever could.