Validation is a soil test before you build the house. Nobody admires soil tests. Everybody regrets skipping them when the walls crack.
Startup validation tests whether people will pay for your solution before you build it. Founders use landing pages, customer interviews, waitlists, fake door tests, and quick MVPs to gather evidence in 30 days or less. The traditional approach builds first, then searches for customers. Most startups fail that way because they spend half a year building something nobody wants. Modern validation flips it. Prove demand exists, then build.
Why does validation matter more than your idea? #
Your startup idea probably is not unique. Someone thought of it before. What matters is execution, timing, and finding the right customers who will actually pay.
As stated by CB Insights in its startup failure research (https://www.cbinsights.com/research/report/startup-failure-reasons-top/), tackling problems that serve no market need is the most cited reason startups fail, showing up in 42% of the classic postmortem analysis. Read that twice. Nearly half of failures are not about money or team or competition. They are about building something nobody wanted badly enough to pay for.
Validation answers five questions. Do people have the problem you think they have. Is the problem painful enough they will pay to solve it. Will they pay you to solve it. How much will they pay. How do they want it solved.
Most founders discover their initial idea was wrong but validation uncovered what customers actually needed. Validation is not about proving you are right. It is about learning what is true. Steve Blank, who gave the startup world customer development, has written for years that founders must get out of the building and test hypotheses with real customers (https://steveblank.com/). The Y Combinator library (https://www.ycombinator.com/library) says the same thing in blunter terms: talk to users, build what they need, repeat.
| Week | Goal | Method | Signal to watch |
|---|---|---|---|
| 1 | Problem proof | 15 to 20 interviews | Repeated pain language |
| 2 | Demand proof | Landing plus waitlist | 5 percent plus conversion |
| 3 | Willingness to pay | Fake door and pre sale | Small paid commitment |
| 4 | Build signal | Thin slice for few users | Use without hand holding |
How do you run week one research and discovery? #
Do not build anything yet. Talk to people. Your goal: understand the problem deeply before you try to solve it.
Days 1 to 2: define your target customer #
Write down exactly who has the problem. Be specific. Solo marketing consultants earning mid five figures annually who juggle multiple client projects is specific. Small business owners is a census category, not a customer.
List where these people hang out online: specific subreddits, Facebook groups, LinkedIn communities, podcasts, Discord servers, Slack communities. Then go there as a learner, not a pitcher. Lurk first. Note the phrases people repeat. Those phrases become your landing page copy later.
Days 3 to 7: conduct 15 to 20 customer interviews #
Reach out in those communities. Ask for 15 minute calls. Do not pitch your solution. Ask about their current workflow and pain points.
Questions that work:
- Walk me through how you currently handle this part of your work
- What frustrates you about your current process
- What have you tried to solve this
- What would make this problem go away
- If there was a tool that solved this, what would it need to do
Then shut up and listen. As Rob Fitzpatrick teaches in The Mom Test (https://www.momtestbook.com/), never ask whether someone would use your idea, because everyone says yes to hypotheticals and nobody means it. Ask about their life, their problems, and their existing behavior. Compliments are worthless. Commitments and stories about past spending are gold.
Take detailed notes. Look for patterns. What do multiple people complain about. What workarounds have they built. What are they already paying for that does not quite work. Ten interviews with consistent pain beat a hundred survey clicks.
How do you run week two landing page validation? #
Build a landing page that describes your solution before the solution exists. See if people care enough to give you their email.
Days 8 to 9: create a landing page #
Use Carrd, Webflow, or BYOB to build a simple one page site in a few hours. Include a clear headline stating the problem you solve, three to five bullets on how it works, benefits instead of features, an email signup form, and no pricing yet. You are testing problem and solution fit, not price sensitivity.
Example headline that works: Stop losing clients to disorganized project updates. Give every client a simple dashboard that shows exactly what is happening with their project.
Example headline that does not work: The revolutionary AI powered project management platform for modern agencies. Nobody wakes up wanting a platform. They wake up wanting the pain gone.
Days 10 to 14: drive traffic to the page #
You need a few hundred visitors for meaningful data. Post in relevant communities with genuine value and mention what you are building. Message the people you interviewed and ask for feedback. Run small Google or Facebook ads in the $50 to $100 range. Share on your own social accounts with context. Ask friends in your target market to share.
Track conversion rate from visitors to email signups. My rough cutoffs from watching founders do this: under 2% means the problem is not painful enough or the messaging is wrong. Two to 5% is decent signal worth refining. Five to 10% is strong. Above 10% means build immediately. These are judgment calls, not physics, so read them alongside interview depth.
A founder testing a freelancer invoicing tool drove 300 visitors to a landing page and got 47 signups, around 15% conversion. Strong signal. Built an MVP in three weeks. Launched to the waitlist. First paying customer within a week. That is what validation success looks like. Fast, specific, paid.
How do you run week 3 fake door testing and pre sales? #
Fake door testing shows people features you are considering and tracks which ones they click before those features exist. It is optional and powerful.
Days 15 to 18: set up the fake door test #
Add a features section to your landing page with four to six potential features. Each links to a page that says the feature is coming soon and invites an email for launch notification. Track which fake features get the most clicks. This tells you what people want most before you build anything.
A productivity app tested five features this way. Calendar integration drew 45 clicks. Time tracking drew 122. Team collaboration drew 23. AI task suggestions drew 89. Mobile app drew 67. They built time tracking first since it carried the strongest signal. Launched with that as the core feature. No debate needed. The clicks voted.
Be honest on the coming soon page. Nobody likes feeling tricked. A short note that you are testing demand before building keeps trust intact.
Days 19 to 21: pre sell if possible #
The strongest validation is money. Try to sell before you build. Put a buy button on your page that leads to a Gumroad or Stripe payment link. When someone tries to buy, tell them the build is finishing, give a realistic timeline, and offer founding pricing for paying now.
If people pay for something that does not exist yet, you have extraordinary validation. Even one pre sale proves someone values it enough to trust you with money. Pre sales fund initial development. They also create accountability, since you promised delivery to a real human.
How do you run week 4 and build a minimal viable product? #
Now you build. Only what is absolutely necessary to test whether people will use it and pay for it.
Days 22 to 25: build the core feature only #
Do not build everything you planned. Build the one thing that solves the core problem. Use BYOB, Bubble, or similar no code tools.
For a client portal: login, project status view, and file uploads. Skip reporting, team management, integrations. For a productivity tool: task creation, basic views, and due dates. Skip automations, calendar sync, mobile app. The MVP should feel incomplete. That is correct. You are testing core value, not building a complete product.
This is where Paul Graham's advice earns its keep. As he writes in Do Things that Don't Scale (http://www.paulgraham.com/ds.html), founders should start with manual, unscalable work to learn what customers need before automating. The concierge MVP, where you run the workflow by hand for five test customers, is not embarrassing. It is the fastest classroom available.
Days 26 to 28: get 5 to 10 users testing #
Go back to your email list from the landing page. Invite the first ten people who signed up. Give them access. Watch them use it over screen recordings or live calls.
Note what confuses them, what they try to do that does not work yet, what they complain about, what makes them say oh, this is useful, and whether they return on day two without prompting. The return visit is the metric that matters. First sessions measure politeness. Second sessions measure value.
Days 29 to 30: decide based on data #
Review all your validation evidence: landing page conversion rate, customer interview patterns, fake door clicks, pre sales, MVP behavior.
Strong validation looks like conversion above 5%, ten plus interviews with consistent pain points, two plus pre sales or strong feature demand, MVP users returning unprompted, and users asking when they can pay.
Weak validation looks like conversion under 2%, interviews revealing happy workarounds, nobody trying to pay, MVP users needing heavy hand holding, and users not returning after the first session.
Strong validation means build. Weak validation means pivot or kill.
What common validation mistakes should you avoid? #
Talking only to friends and family. They will say it is a great idea because they love you. Talk to strangers who have the problem.
Asking would you use this. Everyone says yes to hypotheticals. Ask what they do now, what they tried, what they pay for currently. The Mom Test exists because this mistake is universal (https://www.momtestbook.com/).
Building before validating. The urge to just start building is strong. Resist it. Two weeks of validation saves six months of wasted development.
Ignoring negative signals. If people are not signing up, clicking, or paying, the idea probably does not work. Do not rationalize it away.
Validating with surveys instead of conversations. Surveys have their place but early validation needs real conversations. You learn which questions to ask by talking, not by survey design.
Try it: Site Idea Validator
Which tools help with 30 day validation? #
Landing pages: Carrd, Webflow, or BYOB, with free tiers covering the test. Email collection: ConvertKit or Loops, both with free or cheap starts. Customer interviews: Calendly for scheduling and Loom for recordings. Fake door testing: Hotjar or plain analytics events. MVP building: BYOB, Bubble, or Softr. Payment testing: Gumroad or Stripe, both free plus fees.
Total cost to validate: anywhere from zero to a couple hundred dollars for 30 days. The expensive input is your attention.
What comes next after validation? #
If validation is strong, build the MVP properly over four to eight weeks. Launch to your waitlist. Get the first ten to twenty paying customers. Iterate on feedback. Then scale.
If validation is weak but you learned something, pivot. You talked to real customers. You know a related problem they actually have. Test that.
If validation shows no demand, kill it. Most founders cannot do this. They fell in love with the idea. But killing a bad idea in 30 days beats discovering it after 12 months of building. The soil test did its job.
What are some real validation stories? #
A founder wanted project management for freelancers. Interviewed 25 freelancers. Discovered they already used simple tools and the real pain was client communication and getting paid. Pivoted to a client portal plus invoicing tool. Validated with a landing page near 8% conversion. Built the MVP in three weeks. Hit $2K monthly recurring revenue in 60 days. The interviews built the company, not the original idea.
Another founder wanted a meal planning app for busy parents. The landing page drew 200 visitors and three signups. Interviews surfaced nice idea energy with zero willingness to pay beyond a few dollars, and free apps were good enough. Killed the idea. Saved six months building something with no market.
A third founder wanted a template library for email marketers. Created a landing page with pricing. Drove 400 visitors from marketing communities. Sixty eight people tried to buy. Collected pre sales before building. Built the library in two weeks. Delivered to buyers. Kept selling. The market had voted with cards, not compliments.
What are the trade-offs? #
Thirty days of evidence wins when opinion is cheap. CB Insights ties 42 percent of failures to no market need, so week 1 interviews in Mom Test style, week 2 landing pages against a 6.6 percent median with 5 percent plus as strong, week 3 fake door plus presales, and week 4 concierge MVP in the Paul Graham do things that do not scale spirit prove demand before building.
| Pick the 30 day plan when | Pick a longer path when |
|---|---|
| Willingness to pay can be tested with pages and pilots | The product needs deep tech that demos cannot fake |
| Under 2 percent converts weak and 5 percent plus converts strong | Traffic volume is too thin for rate math and calls must decide |
| Kill, pivot, or build can be called from evidence | Regulation or hardware sets the timeline, not curiosity |
It loses on complex sales and thin samples. Pick the alternative when interviews point to a longer wedge: extend discovery, sell the pilot, then build.
Who this is for (and who should skip it) #
This 30 day plan helps founders testing whether anyone will pay before hiring developers. If you can run interviews, ship a landing page, and invite pre sales in a month, the sequence here makes the call clear.
Skip building weeks until landing and interview signals are strong. If the problem lacks repeated pain language or the waitlist conversion stays flat, keep learning rather than coding.
- Best for founders testing demand with interviews, landing pages, and fake-door offers before building.
- Best for startups running a four-week path from research to waitlist to minimal viable product.
- Best for freelancers helping non-technical clients prove willingness to pay first.
What we learned building this #
BYOB supports the validation loop with browser tools that run without signup, including the site idea validator and project scaffolding from a prompt in the project dashboard. Teams generate a landing page as a SvelteKit app, publish to a preview domain, and measure real clicks before building the thin slice. Live at https://byob.studio/tools/site-idea-validator the validator returns 200 and walks target, pain, and offer checks before any code ships.
Frequently asked questions #
Can I validate a complex technical product in 30 days #
Yes, but you validate the problem and willingness to pay, not the complex tech. Landing pages and customer interviews work for any product. The MVP might be a simplified version or even a manual process you run for five test customers.
What if my idea needs to be secret for competitive reasons #
Ideas are worth little. Execution is. If you are truly worried, use NDAs before interviews. But you will learn faster by talking openly and iterating quickly than by staying stealth.
How many people need to sign up before it is validated #
No magic number. Look for conversion rate plus absolute numbers. In B2B, even ten to twenty highly qualified signups can be enough when each represents budget and urgency.
What if I cannot build the MVP myself #
Use no code tools like BYOB or Bubble. Hire a freelancer for a few hundred to a couple thousand for the simplest possible version. Or run the workflow manually as a concierge MVP where you do everything by hand for five test customers.
Should I charge during validation or give it away free #
Test pricing early. Free users do not validate whether people will pay. Offer a discount for early adopters but charge something. Even a small monthly charge proves more than free signups.
Validate fast. Build when you have proof. Start with BYOB →
