The most expensive mistake in startups
It's not a bad hire. It's not a failed ad campaign. It's not even running out of money.
The most expensive mistake is building something nobody wants.
It happens all the time. A founder spends 6 months and 15,000 euros building a product. They launch it. Silence. No sign-ups. No interest. No revenue. Not because the product was badly built, but because nobody needed it in the first place.
The fix is simple in theory, hard in practice: validate before you build. Here's how to do it properly.
What validation actually means
Validation is not asking your friends if your idea is good. Your friends will say yes because they like you. Your mom will say it's brilliant. Your colleague will say "that's interesting" and never think about it again.
Real validation means getting evidence that strangers would pay money for what you're building. Not "that's a cool idea." Not "I'd definitely use that." Evidence. Behavior. Commitment.
There are levels to this:
- Level 1: Someone says "I have this problem" (problem exists)
- Level 2: Someone says "I'd pay to solve this problem" (willingness to pay)
- Level 3: Someone gives you their email to hear about your solution (soft commitment)
- Level 4: Someone pre-orders or pays before the product exists (hard commitment)
Most founders stop at level 1 and start building. The smart ones push to level 3 or 4 before writing a single line of code.
Step 1: Talk to 20 strangers
Not your friends. Not your family. Strangers who match your target user.
If you're building an invoicing tool for freelancers, go where freelancers hang out. Reddit, Facebook groups, Slack communities, Twitter/X. Post something like:
"I'm researching how freelancers handle invoicing. If you send invoices regularly, I'd love to hear about your workflow. 10-minute call, no pitch, just questions."
Then ask these questions:
- How do you currently handle this? (understand the existing solution)
- What's the most frustrating part? (find the real pain)
- Have you tried other tools? Why did you switch or stop? (understand the competition)
- If something could fix [specific pain], would you pay for it? How much? (test willingness to pay)
20 conversations will teach you more than 6 months of market research. Patterns will emerge. You'll hear the same frustrations repeated. That's your signal.
Step 2: Build a landing page, not a product
You don't need a working product to test demand. You need a page that explains what your product does and asks people to sign up.
A validation landing page has:
- A headline that describes the problem you solve
- Three bullet points explaining how your product works
- A call-to-action: "Join the waitlist" or "Get early access"
- An email capture form
That's it. No product. No backend. No features. Just a page that tests whether people care enough to give you their email.
You can build this in an afternoon with any tool. The point isn't the page itself. It's the data: how many people visit, and how many sign up?
A 10-20% conversion rate (visitors to sign-ups) means strong interest. Under 5% means something is off: either the positioning, the audience, or the idea itself.
Step 3: Drive traffic to the page
Your landing page is live. Now you need eyeballs. You don't need a marketing budget for this.
Free channels that work for validation:
- Post in relevant Reddit communities (don't spam, be genuine, share your story)
- Share on Twitter/X with a clear explanation of the problem you solve
- Post on LinkedIn if your target audience is professionals
- Submit to Indie Hackers, Hacker News, or Product Hunt (Show HN works great for early validation)
- Share in Slack and Discord communities where your target users gather
Aim for 200-500 visitors. That's enough data to know if the idea resonates.
Step 4: Pre-sell before you build
This is where most founders chicken out. But it's the most powerful validation technique: ask people to pay before the product exists.
You can do this with:
- A "Pre-order" button at 50% off the future price
- A "Lifetime deal" for early supporters
- A simple Stripe checkout link for a small amount (9 euros, 19 euros)
If 10 people out of 200 visitors pay you 19 euros, you just made 190 euros and validated that people want your product enough to open their wallet. That's worth more than any survey.
If nobody pays? You just saved yourself months of building something nobody wants. That's also extremely valuable information. Either way, you'll have a much clearer picture when it's time to find your first paying customer.
Step 5: Build only what's validated
You've talked to 20 people. You've tested a landing page. Maybe you've even pre-sold a few spots. Now you know:
- The exact problem people have
- The words they use to describe it (use these in your marketing)
- Whether they're willing to pay and how much
- What features matter most to them
Now build. But only what you've validated. Not the 47 features you dreamed up. The one or two things that solve the core problem.
This is where speed matters. The longer the gap between "someone paid you" and "someone can use the product," the more momentum you lose. Build the MVP in days or weeks, not months.
AI tools make this possible. You can build and launch a startup without coding by describing the product you've validated and letting an AI team build the app, deploy it, and set up everything. Ship fast. Get the product into the hands of the people who already told you they want it.
The signals that tell you to stop
Not every idea validates. That's the whole point of this process. Here are the signals that mean "don't build this":
- Nobody responds to your outreach. You posted in 5 communities and got zero engagement. The problem might not exist, or you're targeting the wrong audience
- People say "that's interesting" but nobody signs up. Polite interest is not demand
- Landing page converts under 3%. The positioning or the idea isn't resonating
- Zero pre-sales after 300+ visitors. If nobody will pay 9 euros for early access, they won't pay 29 euros for the finished product
- Every conversation surfaces a different problem. If there's no pattern, there's no market
Killing an idea early is not failure. It's saving yourself 6 months of building something that was never going to work. Move on to the next idea. The process takes a week, not a year.
The signals that tell you to go all in
On the other hand, here's what "validated" looks like:
- Multiple strangers describe the same problem without prompting
- Your landing page converts above 15%
- People email you asking when the product will be ready
- You have pre-sales before writing a single line of code
- Someone says "I can't believe this doesn't exist yet"
When you see these signals, stop validating and start building. Speed is everything now. Check out the best AI tools to launch a startup in 2026 so you can ship before your early adopters find another solution.
Validation is not optional
It's the single highest-ROI activity you can do as a founder. One week of validation can save you 6 months of wasted effort.
The founders who succeed in 2026 aren't the best builders. They're the best listeners. They talk to users before they build. They test demand before they invest. They let the market tell them what to create.
Your idea might be great. But "might" isn't enough. Test it. Then build it.
Ready to build? See how to build and launch a SaaS with an AI team.