How to Test a Business Idea With $100

A hundred dollars is enough to answer the only question that matters at the start: will a stranger take a real step toward paying you? Spend it on one landing page, one small targeted traffic push, and one direct ask. Everything else, the logo, the business cards, the LLC, is decoration that delays the answer.

What $100 can and cannot buy

It cannot buy proof of a business. It cannot buy a product, a brand, or a market. What it can buy is evidence: a measurable signal that people who do not know you want the thing you are describing.

That distinction matters because founders routinely invert it. They spend thousands building and zero testing, on the logic that the build is the risky part. The build is rarely the risky part. The risky part is that nobody cares, and that is the one risk a hundred dollars can actually retire.

The budget

Here is a sane split:

Landing page: $19. Carrd Pro Lite runs $19 a year and covers a page, a form, and a custom domain. You do not need more.

Domain: about $12. Any registrar. Pick the boring, obvious name.

Traffic: $50 to $60. This is the part founders skip, and it is the entire point. You are buying a small sample of real strangers.

Reserve: the rest, for a second attempt when the first page converts at zero.

Total time: a weekend to set up, one to two weeks to run.

Test one: the fake door

Write the landing page as if the product exists. Headline that names the customer and the outcome, three bullets of specifics, one button: get early access, or better, preorder.

Then put the $50 behind it where your actual customers look. A tightly targeted Meta or Reddit campaign aimed at one niche interest or subreddit, or fifty dollars boosting a post in a community where your buyers already gather. You are not trying to scale. You are trying to find out if anyone outside your head leans forward.

The classic version of this test is older than the tooling. Buffer famously validated with a two-page site before building: page one explained the product, page two said it was not ready and collected emails. Dropbox''s founders have described doing it with an explainer video that sent their waitlist from 5,000 to 75,000 signups overnight. Same principle, smaller budget: describe the thing, measure who acts.

Test two: the deposit

Email signups are weak evidence because they cost the visitor nothing. A dollar amount changes the physics.

Offer early-bird pricing with a small refundable deposit, five or ten dollars, through a simple payment link. State plainly that it is refundable if you do not ship. Ten strangers putting down five dollars tells you more than five hundred signups, because money is the only compliment that cannot be polite.

If deposits feel too strong for your idea, that feeling is data too. It usually means you do not believe strangers would pay, which is worth sitting with before you build.

Test three: ten direct asks

This one is free, and it runs in parallel. Find ten people who match your customer profile, in the communities and threads where they already complain about the problem, and ask for fifteen minutes or a DM exchange. Not "would you use this?" That question begs for a lie. Ask what they do now, what it costs them, and what they last tried.

If you cannot find ten reachable people with the problem, you have learned something the landing page never could: you do not know where your customers are. Fix that before anything else. Our research method for finding where customers hang out walks through it.

Reading the results honestly

After two weeks, sort what you got into three piles.

Money moves: deposits, preorders, direct "how do I buy this" messages. Any nonzero number of these from strangers is a genuine pass signal.

Leans: signups with real email addresses, replies asking follow-up questions, people forwarding your page to someone else. Encouraging, not conclusive.

Politeness: likes, praise, friends sharing it, signups from people who know you. Count these as zero. We wrote a whole piece on why free usage and warm words are not validation.

A total of zero in the first two piles is not a maybe. It is the test working. You just saved yourself six months.

If it fails

Change exactly one variable and run it again with the reserve money. A different customer segment, a different promise, a different problem framing. What you do not change is the test itself, because the test is the only part that is working.

Two clean failures on different framings is a strong answer. At that point the honest move is killing the idea while it still costs you $100 instead of $10,000. If you want a structured way to do that, the free Reality Check will pressure-test the idea against real market evidence and tell you plainly what we tell a lot of founders: sometimes the answer is stop.

Frequently asked questions

Is $100 really enough to test a business idea?

Enough to test demand, not to build the business. A landing page, a small targeted traffic buy, and a direct ask will tell you whether strangers lean forward. What $100 cannot tell you is whether the economics or the product work, and it was never supposed to.

What is a good conversion rate for a smoke test?

There is no universal benchmark, because a niche B2B page and a consumer gadget page behave differently. The honest rule: signups from strangers are encouraging, deposits from strangers are real, and praise from anyone counts as zero.

Should I take preorders for a product that does not exist?

A small, clearly refundable deposit with plain language about the timeline is a standard and honest validation move. Taking money while pretending the product exists is not. The refund promise is what keeps the test clean.

What if my idea needs a physical product?

The test changes shape, not principle. Renderings, a prototype video, or a preorder page all measure the same thing: whether strangers commit. Hardware crowdfunding campaigns are essentially this test with a bigger budget.