In our view, one of the fastest ways to validate a startup idea before you build is to get strangers to give you something costly (their time, their data, a deposit or a signed commitment) for a solution that does not fully exist yet. A common sequence is four tests in order: problem interviews, a landing page test, a pre-sale or letter of intent, and a hand-delivered "concierge" version.
If people won't spend 20 minutes describing the problem, they are unlikely to pay for the fix.
Compliments are free. Commitments cost something. Validation is the work of collecting the second kind. This guide offers criteria, tests and a four week validation sprint you could start on Monday, to adapt to your market.
What startup idea validation actually means
Validation isn't asking friends whether your idea is good. It's running cheap experiments that could prove you wrong, then listening to the answer. This is the core of the lean startup approach Steve Blank describes in Harvard Business Review: test and discard hypotheses with customer feedback before committing to a full product. A validated idea typically has three pieces of evidence behind it:
- A real problem. A specific group of people hit it often, it costs them something measurable, and they have already tried to solve it.
- A reachable customer. You can find and contact these people repeatedly without a big marketing budget.
- Willingness to pay or switch. Some of them will commit money, time or a public promise before the product is finished.
Notice that "the product works" is not on the list. Building comes after. If you are still choosing what to test, a list of startup ideas for teens can offer low-cost starting points.
Why you should validate a startup idea before writing code
The real cost of a wrong idea is rarely the code. It's a year of your life, the co-founder relationship, and the credibility you spend with early customers and investors.
- Missing demand often kills companies. CB Insights' March 2026 analysis of 431 venture-backed startups that shut down since 2023 found that, among the 385 with identifiable reasons, 43 percent cited poor product-market fit. Running out of capital topped the list at 70 percent, but the report calls it almost always the final cause, not the root problem.
- Investors tend to fund evidence. Carta describes a typical pre-seed round as roughly $250K to $1M, enough for at least a year to a year and a half of runway, and its State of Pre-Seed report for Q2 2026 put the average SAFE or convertible note at $276K, the highest in more than four years. Few investors write that check for an idea alone; the investor-side pre-seed funding guide shows what early backers look for.
- Validation is cheap by comparison. A sprint like the one below usually costs a few hundred to a couple of thousand dollars in tools and ads, plus several weeks of founder time. Building the wrong product can cost months of salary or contractor fees even on a lean budget.
- Conversations tend to change ideas. Twenty honest conversations with potential customers often change something important: the segment, the price or the problem itself. Cheap changes early usually beat expensive ones later.
This is the gap 1752vc's Launchpad is built to close. It is a 12-week, self-paced, remote sprint that takes aspiring founders from -1 to 1: validate an idea, find a first customer and build a path to traction. For founders at this stage, we think validating before building is most of the game.
How to validate a startup idea step by step
Step 1: Name the assumption that would kill the idea
Every idea rests on a few beliefs. List them and rank them by how much damage it would do if each were false. A simple illustrative template:
- Who: independent physical therapy clinics with 1 to 5 staff.
- Problem: they spend 6 or more hours a week chasing insurance pre-authorizations.
- Current fix: a front-desk employee and a spreadsheet.
- Our bet: they would pay $150 to $300 a month to cut that to 1 hour.
- Riskiest assumption: the pain is bad enough to change their workflow.
Test the riskiest assumption first. If the problem isn't painful, little else matters. We'd much rather back a painkiller than a vitamin (our take on vitamins vs painkillers).
Step 2: Problem interviews (week 1)
Problem interviews are 20 to 30 minute conversations with people who match your "who." The aim is not to pitch but to listen for how the problem shows up in their week. Rob Fitzpatrick's book The Mom Test makes the core point well: ask about specific past behavior, not opinions about your idea, because people will be polite about ideas. Our customer interview guide has a full script and the questions to avoid.
You can often find 15 to 20 people through your network one step removed, communities where your customer already gathers, and honest cold outreach that says you are researching, not selling. Five questions we find useful:
- Walk me through the last time this problem came up.
- What did you do about it?
- What did that cost you in time, money or stress?
- Have you tried other solutions? What happened?
- If this went away tomorrow, what would change for you?
Good answers include specific incidents, numbers and a workaround they already pay for. Weak answers sound like "yeah, that would be nice." Politeness is noise. Effort is signal. Log each interview (segment, pain score from 1 to 5, current fix, one quote). After 15 conversations you will often see whether pain clusters in one segment. That segment is a natural beachhead.
Step 3: The landing page test (week 2)
A landing page test, often called a smoke test, shows whether strangers respond to the promise before the product exists. A typical setup is a one-page site (a few hours with no-code tools) with a headline that names the customer and the outcome, three short benefits, one call to action, and a form with one qualifying question. The page needs a name and a domain first, and the guide to naming your startup keeps that decision to a week rather than a month.
As a rough guide, 300 to 1,000 targeted visitors is a reasonable sample. A few hundred dollars of paid search or social ads is usually enough for a signal, and posting in the communities you used for interviews is free and often better.
For context, Unbounce's 2024 Conversion Benchmark Report, based on more than 41,000 landing pages, found a median conversion rate of 6.6 percent across industries, with individual industries ranging from 3.8 to 12.3 percent. It is a reference point, not a target; the rules of thumb below vary with what you ask for:
- Waitlist signup: a focused B2B page to a well-targeted audience may beat the cross-industry median; broad consumer traffic often converts lower.
- "Book a call": often a small fraction of your waitlist rate, because the ask is bigger.
- Paid deposit or pre-order: even low single-digit percentages are a meaningful signal, because money changed hands.
Very low conversion on well-targeted traffic often suggests the message or the market is off. Many founders revisit the message first, then question the market.
Step 4: Pre-sales, deposits and letters of intent (week 3)
This is where many ideas stall, and in our view it is the most valuable week of the sprint. We put more weight on a pre-payment than on any number of enthusiastic interviews (why we think opinion-only discovery falls short). Ask people to commit before the product exists:
- Consumers: a refundable deposit or a paid pre-order.
- Small businesses: a signed pilot agreement with a price, even if it starts in 60 days.
- Enterprise: a non-binding letter of intent naming the budget owner, use case and target start date.
Show your best 10 interviewees a clickable mockup and ask: "If this existed in 8 weeks at $X, would you sign up for a pilot today?"
Then stop talking. The silence is part of the test. As a rule of thumb, a reasonable B2B target is 3 to 5 signed pilots or LOIs from 20 asks; for consumer products, dozens of deposits from a few hundred waitlist names.
Step 5: The concierge MVP (week 4 and beyond)
A concierge MVP delivers the outcome by hand for a few paying customers before you automate anything. If the promise is a one-hour pre-authorization process, you personally handle pre-authorizations for five clinics.
You learn the real workflow, which parts customers care about, and the true cost to serve, while earning revenue before hiring an engineer. A typical run is 4 to 8 weeks with 3 to 10 customers, then automating only the step that eats most of your time. That first automated step is arguably your real MVP. If you are building without engineers, the no-code for startups guide shows how far you can get.
A sample 4 week validation sprint checklist
Week 1: Problem. Write the assumption sheet. Complete 15 to 20 interviews. Find the segment where pain averages 4 or higher.
Week 2: Message. Launch a one-page site with a single call to action. Send 300 to 1,000 targeted visitors. Test three headlines and keep the winner.
Week 3: Commitment. Build a clickable mockup. Make 20 direct asks for deposits, pilots or LOIs. Record every objection word for word.
Week 4: Delivery. Onboard 3 to 5 concierge customers. Track hours per customer and the outcome delivered. Decide: proceed, change segment, or stop.
Decision criteria: go, pivot or stop
Set thresholds before you start. Afterward, you'll be tempted to grade on a curve. These are our suggested rules of thumb for a B2B idea; it is worth adjusting them to your market before you begin:
- Go: at least 60 percent of interviewees in one segment rate the pain 4 or 5, landing page conversion beats the 6.6 percent cross-industry median, and you have 3 or more paid pilots or LOIs.
- Pivot: the pain is real but spread across segments, or people love the problem statement but will not commit money. Consider rerunning weeks 1 to 3 with a narrower customer; our guide on how to pivot a startup covers that decision in more depth.
- Stop: fewer than 30 percent report meaningful pain and conversion stays very low after three headline tests. That's a good outcome. It cost you a month, not a year.
Many investors apply a similar filter. The investor-side guide on how venture capitalists make investment decisions shows how much weight they put on evidence of demand at the earliest stages.
"But nobody asked for the iPhone"
A classic objection. Some breakthrough products had no obvious demand before they existed, and customers can't always describe what they'll love. Run every idea through a four week demand test and you might kill something new.
But.
Most startups aren't building a new category, and the ones that are still get tested, just later and more expensively. Even a truly new product solves an old problem: people wanted to communicate, find things, get somewhere. Validating the problem is still possible when the solution is hard to imagine. It's the solution you hold loosely.
Where we land
Validate the problem hard, the solution lightly, and the willingness to pay before you write much code. If the evidence says stop, a month is a cheap price for the answer.
That's our bias. Deep tech and research-heavy ideas may need a different test, because the risk sits in whether it can be built at all.
Common mistakes when validating a startup idea
- Interviewing friends and family. They are often too kind to be candid, and they are rarely your customer.
- Pitching during problem interviews. Once you describe the solution, people tend to start being kind instead of honest.
- Treating signups as validation. Signups show interest. Deposits and signed pilots are much stronger evidence.
- Testing a vague segment. "Small businesses" is too broad to be a useful segment. "Independent dental practices in Texas with 2 to 4 chairs" is.
- Skipping week 3. Asking for money is uncomfortable. That discomfort is part of the test.
Once you have evidence, the next question is whether there is a business behind it. The guide on how to find product-market fit picks up where validation ends, and the co-founder playbook covers who you might be doing this with.
The bottom line
Validation doesn't prove an idea will work. It tells you, cheaply and early, whether anyone will pay you to find out.
Interest is a signal.
A deposit is a decision.
Key takeaways
- In our view, the stronger validation comes from strangers committing time, data or money before the product exists, not from collecting opinions.
- Testing the riskiest assumption first helps; often that is whether the problem hurts enough to change behavior.
- CB Insights' 2026 shutdown analysis found 43 percent of failed VC-backed startups with known reasons cited poor product-market fit, which is what validation is meant to catch early.
- One sequence is problem interviews, a landing page test (benchmark: Unbounce's 6.6 percent cross-industry median), a pre-sale and a concierge MVP over four weeks.
- Setting go, pivot and stop thresholds before you start lets the data carry more weight than your attachment to the idea.
Frequently asked questions
Problem interviews cost only time, and community posts can drive landing page traffic for free. Many no-code site builders have free or low-cost plans. In our view, the hard part of validation is your willingness to ask people for a real commitment, not the tools.
As a rule of thumb, 15 to 20 conversations within a single, narrow customer segment is enough. Patterns often become clear somewhere between interview 10 and 15. If you are still hearing brand-new problems at interview 20, your segment may be too broad. It often helps to narrow it and run another round rather than averaging mixed signals.
Unbounce's 2024 benchmark report found a median of 6.6 percent across more than 41,000 landing pages, with industry medians from 3.8 to 12.3 percent. A focused page aimed at a well-defined audience may beat the median for a free signup, while paid deposits convert much lower and can still count as a strong signal.
A concierge MVP is a version of your product where you deliver the outcome by hand for a few paying customers before automating anything. It tests whether people will pay, teaches you the real workflow and shows which step to automate first.
Usually not, in our view. A waitlist shows interest, which is cheap to give. Stronger validation tends to need a costly signal: a deposit, a signed pilot, a letter of intent or repeated use of a manual version. One way to use the waitlist is as a list of people to ask for a commitment.
Sources
Disclaimer: This guide is for general education only and is not legal, tax or investment advice. Laws, market data and program terms change, so it may not reflect the latest developments or fit your situation. Treat it as a starting point, not a source of truth, and talk to a qualified lawyer, accountant or financial adviser before you make decisions.
