Kill Tests: 12 Cheap Experiments That Disprove an Idea in a Week

Most advice about how to test a business idea cheaply quietly turns into "how to collect compliments cheaply." This list is different. Every experiment here is designed to kill the idea if it deserves killing, in seven days or less, for $100 or less. If your idea survives all the ones you throw at it, you've earned the right to build.

What a kill test is

A kill test is an experiment built to disprove your business idea, not to confirm it. You define the failure condition in advance ("fewer than 2 yeses out of 20 asks kills this"), you run it against strangers who actually have the problem, and you obey the result. It matters because the top startup killer, no market need, shows up in 42% of CB Insights' failure postmortems, and it's detectable in a week if you ask the market directly.

The 12 kill tests

1. The direct ask ($0). Message 20 people who have the problem. Describe the offer, name the price, ask them to buy. Under 2 yeses is a kill.
2. The pre-order page ($0 to $30). One plain page: the promise, the price, a buy button that leads to a deposit or payment form. Real intent only.
3. The $50 ad test ($50). Run $50 of tightly targeted ads to that page. You're buying click and signup data from the actual market, not impressions.
4. The fake door ($0). Add the button or menu item for the thing before it exists and count clicks. Clicks are curiosity; pair this with an ask before trusting it.
5. The concierge week ($0). Deliver the service by hand to 3 paying customers. If they won't pay for the handmade version, the software version won't save you.
6. The competitor complaint hunt ($0). Spend an hour in 1 star reviews and complaint threads about existing solutions. No complaints worth solving means no gap worth filling.
7. The cold call ten ($0). Call 10 target businesses. You're counting booked meetings, not pleasant chats. Zero meetings from ten calls is loud.
8. The price doubling test ($0). Quote double your planned price to the next 5 prospects. If everyone accepts without flinching, you underpriced. If everyone walks, the offer wasn't wanted at any price.
9. The deposit swap ($0). Replace "join the waitlist" with "reserve your spot for $10." Watch signups convert to silence or to customers.
10. The marketplace listing ($0 to $20). List the product where buyers already shop: Etsy, eBay, Fiverr, Upwork, a niche directory. Existing demand pools are the cheapest test traffic on earth.
11. The interview to invoice test ($0). End every customer interview with "can I invoice you for the first month?" The pause before the answer tells you everything.
12. The 7 day content test ($0). Post daily for a week in the community where your customers live, solving the problem in public. Buyers show up in your messages asking the price. Nobody asking is an answer too.

How to read the results without lying to yourself

Set the kill line before you start, in writing, while you're calm. "I'll know more after I tweak it" is how a one week test becomes a six month hobby. And count the right currency: money first, then time (booked calls, signed pilots), then reputation (introductions). Likes, upvotes, and "this is so cool" are not on the list. If you're unsure whether your idea is bad or just untested, these bad idea signals will help you tell the difference.

What to do when an idea survives

Survival is not a build order, it's a license to run the next, slightly more expensive test. Let the evidence grow with the stakes: interviews, then deposits, then a concierge version, then the smallest sellable build. If you want help pressure-testing what survived, the Reality Check runs the demand, competition, and risk evidence, and the Opportunity Finder can hunt for the sharpest version of the idea before you commit.

The Marketur take

A failed kill test is the cheapest education in business. You paid $50 and a week to learn what the optimistic founder pays $15,000 and a year to learn. Run the tests. Obey the results.

Frequently asked questions

What is a kill test for a business idea?

A kill test is a small, fast experiment designed to disprove an idea while it's cheap to be wrong. You set the failure condition in advance, run it against real potential buyers, and accept the verdict.

How much should testing a business idea cost?

Most ideas can be meaningfully tested for $0 to $100. If a test needs thousands of dollars, it's not a test anymore; it's a launch. Break it into cheaper questions.

What if my kill test fails?

You just saved months. Write down what the market told you, keep the customer insight, and either pivot the solution or move to the next idea with the same discipline.

Can I test a business idea with no money at all?

Yes. Ten of the twelve tests above are free. The direct ask, cold calls, deposit swaps, and marketplace listings all cost time instead of money, and they're usually the most honest tests on the list.