Last year, a founder I was advising spent €15,000 building an app before talking to a single potential customer. When she finally launched, the silence was deafening. Not because the app was bad — it was actually well-built — but because the problem she was solving didn’t exist the way she imagined it.
I’ve seen this story play out hundreds of times across 20+ years of working with startups. The pattern is always the same: someone gets an idea, gets excited, starts building, and then discovers — too late — that nobody wants what they made.
The fix is embarrassingly simple. Test first. Build later. And you don’t need a single euro to do it.
The Free Validation Stack
There’s a stack of validation methods that cost nothing but your time and willingness to be uncomfortable. I call them “free” not because they’re easy, but because they don’t require capital.
Here’s what the stack looks like:
-
Conversation mining — Go where your potential customers already talk (Reddit, forums, LinkedIn groups, Facebook communities) and read what they complain about. Don’t post. Don’t pitch. Just read.
-
Direct outreach — Send 20 messages to people who fit your customer profile. Ask them about the problem you think exists. Not your solution. The problem.
-
Landing page test — Create a one-page site describing what you plan to build. Add an email signup. Drive traffic through organic posts in communities you’ve already been reading.
-
Pre-sell attempt — Before building anything, try to get someone to pay for it. A simple PayPal link or Stripe checkout page is enough.
-
Manual delivery — If someone pays, deliver the service manually. No automation. No product. Just you, doing the thing by hand.
Each layer builds on the previous one. You don’t move to layer two until layer one gives you signal. And most ideas die at layer one or two — which is exactly what you want. Killing bad ideas early is the entire point.
The biggest mistake I see is founders skipping to layer three or four without doing the first two. A landing page without conversation mining is just guessing with better graphics.
Conversation Mining: The Underrated Superpower
Here’s a specific technique I use. I call it the “complaint catalog.”
Open a spreadsheet. Three columns: Platform, Complaint, Frequency. Then spend three hours reading through forums, subreddits, and community groups where your target customer hangs out.
Every time someone describes a frustration related to your space, log it. Don’t interpret. Don’t filter. Just capture the exact words they use.
After three hours, sort by frequency. The complaints that show up again and again — in different communities, from different people, using similar language — those are real problems. Everything else is noise.
When I was helping build the Startup Burgenland accelerator, we used this method to identify which support services founders actually needed versus what we assumed they needed. The gap was massive. We thought founders wanted mentorship. What they actually wanted was someone to tell them their idea was bad before they wasted six months on it.
This method works because people complain honestly in communities. They don’t complain honestly to your face when you ask “would you use this?” That question is broken, and I’ll get into why in another section.
The raw material you collect here becomes your messaging later. When you eventually write a sales page or landing page, you’ll use the exact language your customers use to describe their own pain. This is not clever marketing. It’s just listening.
The 20-Message Test
Once you’ve got a complaint catalog, pick the top three complaints. Now it’s time to talk to humans.
Send 20 messages. Not 5 — that’s too few to see patterns. Not 50 — that’s procrastination disguised as research. Twenty is the sweet spot.
Who do you message? People who match your target customer and who you can find on LinkedIn, Twitter, or through mutual connections. The message is simple:
“Hey [name], I’m researching [specific problem area] and noticed you [specific thing that makes them relevant]. Would you be open to a 15-minute call? I’m not selling anything — genuinely just trying to understand the problem better.”
That last line matters. You need to mean it. If you go into these calls with a pitch, you’ll get polite lies. If you go in genuinely curious, you’ll get truth.
From 20 messages, expect about 5-7 replies. From those, expect 3-4 calls. From those calls, you need to hear whether the problem you identified is real enough that people have already tried to solve it, are currently spending money on bad solutions, or are actively frustrated.
If all four people say “yeah, that’s annoying but I wouldn’t pay to fix it,” you have your answer. Move on. If even two say “I’ve been looking for something like this and can’t find it,” you have signal.
This entire process costs zero euros and takes about a week. Compare that to the €15,000 app build I mentioned earlier. Starting now with imperfect information beats starting later with a fully built product nobody wants.
Why “Would You Use This?” Is a Broken Question
Let me save you from the most common validation mistake.
Never ask: “Would you use this?” or “Would you pay for this?” or “What do you think of this idea?”
These questions produce garbage data. Here’s why: humans are polite. When someone sits across from you (or on a Zoom call) and you describe your baby — your precious idea — they’ll smile and say “yeah, that sounds interesting.” They don’t want to hurt your feelings. They don’t want the social awkwardness of saying “no, that’s dumb.”
So they lie. Not maliciously. Just… reflexively.
Instead, ask about behavior. Past behavior, specifically.
- “When did you last encounter this problem?”
- “What did you do about it?”
- “How much did that solution cost you?”
- “What’s still not working about how you handle this?”
These questions are hard to fake. Someone either has a story or they don’t. If they struggle to remember the last time they faced the problem, the problem isn’t urgent enough. If they describe spending real money on a bad solution, you’ve struck something worth pursuing.
This approach comes from Rob Fitzpatrick’s work on customer conversations, and it’s been the single most valuable validation tool in my consulting practice. I’ve used it with 40+ startups at Startup Burgenland and it consistently separates real opportunities from founder fantasies.
The rule is simple: never talk about your idea. Only talk about their life. Your idea is irrelevant at this stage. Their behavior is everything.
The Landing Page That Tests Truth
Okay — let’s say your conversations confirmed the problem is real and people are spending money on bad solutions. Now you can build a landing page. Still costs nothing if you use free tiers of tools like Carrd, Notion, or even Google Sites.
But here’s the critical part: your landing page isn’t a brochure. It’s a test.
The page should do exactly three things:
- Describe the problem — using the exact language from your complaint catalog
- Promise a specific outcome — not features, but the result the customer gets
- Have one call to action — email signup, waitlist, or ideally a pre-order button
Then share it. Not with friends and family (they’ll sign up to be supportive, which gives you false signal). Share it in the communities where you did your conversation mining. Share it with the people you interviewed. Post it where your target customers actually spend time.
Track two numbers: click-through rate and conversion rate. If people click but don’t sign up, your copy is wrong. If people don’t click at all, your offer isn’t compelling. If people sign up and reply to your confirmation email asking when the thing launches, you’ve got real interest.
I’ve seen founders treat landing pages like art projects — spending weeks on design, agonizing over color choices. That’s the preparation trap. Your landing page should take two hours max. It should be ugly. It should have typos. The only thing that matters is whether people give you their email address.
Pre-Selling: The Ultimate Validation
Here’s where most people chicken out: asking for money before the product exists.
I get it. It feels wrong. It feels presumptuous. It feels like you’re scamming people by selling something that doesn’t exist yet.
But pre-selling is the single most honest form of validation. Because money is the only signal that doesn’t lie.
Email signups can be casual interest. Survey responses can be polite agreement. But when someone pulls out their credit card and pays you actual money for something that doesn’t exist yet — that person has a real problem they want solved.
The mechanics are simple. Set up a Stripe payment link. Price your thing at whatever feels reasonable (we can refine pricing later). Send it to your waitlist or to people from your interviews.
The message: “I’m building [thing that solves problem]. It’ll be ready in [timeframe]. If you pre-order now, you get [early bird price / bonus / founding member status]. If it doesn’t ship or you’re not happy, full refund, no questions.”
If nobody buys, you just saved yourself months of building the wrong thing. If even 3-5 people buy, you have real validation and your first customers.
When I exited Vulpine Creations, one of the things I’d learned was that revenue is the only metric that matters for validation. Everything else is a proxy. Money is the signal.
Putting It All Together: A 14-Day Validation Sprint
Here’s the exact timeline I recommend:
Days 1-3: Conversation mining. Build your complaint catalog. Identify the top three problems.
Days 4-7: Send 20 messages. Do 3-4 customer interviews. Refine your understanding of the problem.
Days 8-9: Build a landing page. Describe the problem and the promised outcome.
Days 10-12: Share the landing page. Drive traffic. Track signups.
Days 13-14: Email your signups with a pre-sell offer. Track purchases.
Two weeks. Zero euros. And at the end, you’ll know more about your market than most founders know after spending €10,000+ on development.
The hardest part isn’t any individual step. The hardest part is accepting that your idea might be wrong. That the problem you’re passionate about solving might not be a problem anyone else cares about. That’s ego work, not business work. But it’s the most important work you’ll do.
I’ve watched founders fall in love with solutions. The good ones fall in love with problems instead. And problems are free to study.
Key Takeaways
- Validation costs time, not money. The five-layer free validation stack (conversation mining, direct outreach, landing page, pre-sell, manual delivery) covers everything you need.
- Never ask “would you use this?” Ask about past behavior instead. People lie about intentions but tell the truth about what they’ve already done.
- Money is the only signal that doesn’t lie. Pre-selling before building is the most honest validation method available.
- Two weeks is enough. A structured 14-day sprint gives you more market intelligence than months of building in silence.
- Kill bad ideas early. The goal of validation isn’t to confirm your idea. It’s to find out if the idea should die before you invest real resources.
