Validate Your Idea Before You Build It

You have the idea. You can already see the architecture in your head, the clean little dashboard, the pricing page with three tiers. The urge to open your editor and start is almost physical. That urge is the most dangerous moment in a solo builder's career, because building is the one part of this you are good at, and it feels like progress even when it is quietly walking you off a cliff.

Here is the pattern that gets so many capable people. You spend two months building something genuinely good. You launch. Nothing happens. Not rejection, just silence, which is worse. And the reason is almost never the code. The reason is that you never actually confirmed anyone wanted the thing before you built it. You validated your own excitement, which is easy, instead of validating demand, which is the only thing that pays.

Where does the money actually come from?#

Money shows up when someone with a painful, recurring problem decides your solution is worth paying for instead of continuing to suffer or continuing to pay for whatever they use now. Every dollar traces back to that decision. Validation is just moving that decision earlier, before you have spent anything, so a "no" costs you a coffee instead of a quarter of your year.

A real problem someone already pays to avoid
        |
        v
  You find the people who have it (a narrow segment)
        |
        v
  You talk to them and confirm the pain is real and expensive
        |
        v
  You ask for a buying signal (pre-order, deposit, "take my card")
        |
        +--> polite interest, no signal --> STOP, rework or move on --> $0 spent building
        |
        v
  Real signal (money or a firm commitment)
        |
        v
  NOW you build, with buyers already waiting
        |
        v
  You ship to people who asked for it --> revenue

Notice where the "stop" branch is. It is before you build, not after. The entire point is to catch the bad ideas at the cheap stage. A validated idea does not guarantee success, but an unvalidated one is a coin flip you paid for with months of work. If the deeper economics of how online products actually generate revenue feel fuzzy, where does online money come from walks through the mechanics.

How validation actually works#

Validation is not a survey and it is not asking your friends. It is a short sequence of real-world tests that get progressively more expensive to fake, which is exactly why they work.

Start with a problem, not a solution. The strongest ideas come from a problem people already spend money or serious effort to avoid. If a business is currently paying a virtual assistant to do something by hand, paying for three tools that half-solve it, or losing money every month because of it, that is a validated problem before you build anything. The spending is the proof. Your job is to find where money or painful workarounds already flow, because that is where a budget already exists.

Talk to the people who have the problem. Not to pitch. To understand. You want to hear them describe the problem in their own words, tell you what they currently do about it, and mention what that costs them in money or time. When someone says "I've built a messy spreadsheet to handle this and it breaks every month," you have found gold. That sentence is worth more than a hundred survey responses. We cover how to run these conversations without leading the witness in how to talk to users.

Narrow your target hard. "Small businesses" is not a segment, it is a fantasy. "Bookkeepers who serve dental practices" is a segment you can actually find, talk to, and sell to. A narrow segment makes every later step easier, because you can go to the exact place these people already gather and say something that sounds like it was written for them. Pick a narrow first segment makes the case for going smaller than feels comfortable.

Then push for a buying signal. This is the step everyone skips, and it is the only one that counts. Interest is free. A signal costs the other person something: money, a deposit, a signed commitment, a spot on a waitlist they had to enter a card to hold. The moment you ask for something real is the moment polite people stop nodding and honest data appears.

The "would you pay" test versus polite interest#

Here is the trap. You describe your idea, and people say "oh that's a great idea, I'd totally use that." You walk away glowing. That sentence is worthless. People are wired to be nice to you, especially to your face, and "I'd use that" costs them nothing to say.

The test is not "would you use it." The test is "would you pay for it, how much, and can I take a deposit today." Watch what happens to the energy in the room when you ask that. Polite interest evaporates. You find out fast whether you were talking to a buyer or to someone being kind.

Consider a clearly hypothetical example to see the difference. Say you are thinking about a tool that auto-generates compliance reports for the dental bookkeepers above. You talk to ten of them. All ten say the reports are a pain. Nine say "yeah I'd use something like that." Good feeling, useless data. Then you say: "I'm building it now. It'll be about $40 a month (this number is hypothetical, just to test the reaction). I'm taking $20 pre-orders that lock in your first three months at that price. Want in?" Suddenly, in this made-up scenario, two of the nine reach for a card and seven suddenly remember reasons to wait. Those two are your business. The seven were being polite.

The number matters less than the reaction to it. You are not running the transaction to profit, you are running it to separate buyers from bystanders. Money changes what people tell you, because now the answer costs them something. For a deeper look at reading these signals, is the demand real: interest versus buying intent is the companion to this whole idea.

The fake-door and pre-sell test#

You do not need the product to test whether people will buy it. You need a page that describes it clearly and an ask. This is the cheapest, most honest test available to a solo builder.

Build a simple landing page. Describe the specific problem, the specific solution, and a price. Then put a button on it. The button can do one of two things, and which you choose depends on how honest you want to be with early visitors:

A pre-sell takes real money now for the product you are about to build, usually at a discount, with a clear promise about when it ships and a no-questions refund if you do not deliver. This is the strongest signal that exists. Someone handing you money for something that does not yet exist is not being polite. They are voting with the only vote that counts.

A fake-door test puts up the button, and when someone clicks "Buy" or "Start free trial," they land on a page that says the product is launching soon and asks for their email to be first in line. You are measuring how many people click the real buying button, not just how many read the page. Be careful here: never take money for something that does not exist and never intend to build, and make the "coming soon" honest the instant they click. The goal is a demand measurement, not a bait-and-switch.

Either way, the metric you care about is the drop from "visited" to "took a real action." A page that a hundred people read where zero click buy is telling you something loud. A page where several people try to pay is telling you to open your editor. You can drive a small amount of traffic to this page from the exact community where your narrow segment gathers, which is the same skill you will need later anyway. Get your first 10 customers covers how to find and reach those first real people once you are ready to sell for real.

A clearly hypothetical worked example#

Let me put fake numbers on the whole flow so the shape is clear. Every figure here is invented to illustrate the method, not a claim about results.

Say you spend one week talking to fifteen people in your narrow segment. Twelve confirm the problem is real and annoying. You spend a day building a one-page pre-sell site describing the tool at a hypothetical $30 per month, with a $15 founding-member pre-order. You spend the next week posting helpfully in the two communities where these people already gather, and you send the page to the twelve you spoke with.

In this made-up scenario, 80 people visit the page. 22 click the buy button. 6 actually complete the $15 pre-order. That last number, 6 paying strangers before a single feature exists, is a genuinely strong signal for a niche B2B tool. If instead 80 people visited, 3 clicked buy, and 0 paid, you just saved yourself three months of building. Both outcomes are wins, because both cost you about two weeks instead of a quarter of a year. That asymmetry is the entire reason validation exists.

What you need and what it costs#

Required:

• A landing page. A no-code builder or a single static page is fine. Cost is roughly $0 to $20 for the month.
• A way to collect money or a signal. A payment link from a standard processor works and costs nothing until you take a payment.
• Access to your segment. The communities, forums, or professional groups where they already are. Free, but it costs your time to show up honestly.
• Your own time for conversations. Ten to twenty short calls or chats. This is the real cost, and it is unavoidable.

Optional:

• A custom domain, maybe $10 to $15 for the year, mostly for looking legitimate.
• A small ad budget to drive extra traffic to the fake-door page if your community reach is thin. Keep this tiny, since the point is a signal, not scale.
• An email tool to collect the waitlist. Most have a free tier that is plenty at this stage.

The whole test should cost you tens of dollars, not hundreds. If you are being tempted into a $500 stack of tools to run a validation test, you have already lost the plot.

How long it takes#

Plan for one to three weeks, not months. Roughly: a few days to define the problem and segment, a week of conversations, a day or two to build the page, and a week to drive a little traffic and count the signal. If you cannot get a clear read in three weeks, that itself is data. It usually means the segment is too vague or you have not actually found where these people gather. Tighten the segment and try again rather than pushing forward on hope.

What beginners get wrong#

They validate their excitement instead of demand. They ask "is this a good idea" (everyone says yes) instead of "will you pay for this" (the truth appears). They target "everyone," which means they can find no one. They treat a big waitlist of free emails as proof, when a waitlist that never had to pay anything is mostly curiosity. They ask leading questions that beg for a yes. And the big one: they start building anyway, because building feels safe and asking strangers for money feels scary. The fear is the whole reason it works. If it were comfortable, it would not filter anything.

How I would start#

Here is the exact sequence I would run this week, in order:

1. Write down the single most specific version of the problem you can, and the exact type of person who has it. Not "freelancers," but "freelance video editors who bill hourly."
2. Find the two places those people already gather online. A subreddit, a Slack, a forum, a Facebook group. Join and read for a day before saying anything.
3. Reach out to ten of them for a fifteen-minute conversation. Ask what they currently do about the problem and what it costs them. Do not pitch. Listen.
4. If most confirm the pain and mention real cost or ugly workarounds, build a one-page pre-sell site over the next day. Clear problem, clear solution, a hypothetical price, a real buy button.
5. Put a small pre-order or paid founding-member offer behind that button, with an honest refund promise.
6. Send the page to the ten you spoke with, and post helpfully (not spammily) in the two communities.
7. Count real actions, not compliments. If several people try to pay, open your editor. If nobody does, go back to step one with what you learned.

What I would not do#

I would not build a working product to "see if people like it." That is the expensive test disguised as the cheap one. I would not run a survey and treat the results as demand, because surveys measure opinions and opinions are free. I would not target a broad market to keep my options open, because broad means unreachable. I would not take real money for something I have no intention of building. And I would not let a pile of free waitlist signups convince me I have a business, because until a card comes out, interest is just weather.

Close#

The cheapest version of your idea is the one you never build. Validation is how you find out which ideas those are, before they cost you a season of your life. It is not glamorous and it does not feel like real work, because real work, to a builder, feels like typing. But the founders who make it a habit stop wasting months on things nobody wanted, and they start every build with buyers already waiting.

Run the test. If people pay, you have permission to build the thing you were itching to build anyway, now with the confidence that someone is on the other end. If they do not, you just bought that confidence for the price of a couple of weeks. Either way you win. When the signal is real, get your first 10 customers is where you go next.