How to find out if a business idea sells in 30 days, with real buyers
Put a real product with a real price in front of real buyers for 30 days and count visitors, requests, calls and payments. The bar: 300, 10, 3 and 1.
You find out if a business idea sells by putting a real product with a real price in front of real buyers for 30 days and counting four numbers: visitors, requests, calls and payments. A fair first bar, and the one we work to, is 300 visitors, 10 requests, 3 calls and 1 paying customer; clear it and you continue, miss it by a wide margin and the pitch, the price or the audience has to change.
Surveys and friends cannot answer this question. Only a checkout can. CB Insights' 2026 analysis of 431 venture-backed companies that shut down found that 43% cited poor product-market fit and 70% ran out of capital, usually after building for far longer than 30 days.
Which four numbers count, and what is normal?
| Step | What counts | Published benchmark | 30-day bar |
|---|---|---|---|
| Visitors | People who reach your page from search ads, direct messages or your own audience | Search ads: 6.64% click-through rate and USD 5.42 average cost per click across industries, USD 5.87 in business services (LocaliQ, 2026) | 300 |
| Requests | A form sent, a trial started or a checkout opened | Landing pages convert 6.6% at the median, 6.1% in professional services, and 75% of pages get at least 3.6% from Google search ads (Unbounce, 2024) | 10 |
| Calls | A conversation with a buyer, booked or taken | Cold email replies average 3.43%, with the top quarter above 5.5% (Instantly, 2025 data); lists under 50 contacts get 5.8% against 2.1% for lists of 1,000 (Belkins data via Woodpecker) | 3 |
| Payments | Money moved to your account | Search ads turn 8.18% of clicks into leads across industries and 4.85% in business services, at USD 93.69 per lead (LocaliQ, 2026) | 1 |
The 10, 3 and 1 are the verdict numbers on cofounderdesk.com; the exact bar moves with the price and the product, and it is written down before the first visitor arrives.
What do 10 requests, 3 calls and 1 payment tell you?
Ten requests from 300 visitors is 3.3%. That is under the 6.1% median for professional services but close to the 3.6% that three quarters of search-ad pages clear, so the page is understood and the offer is not absurd. Three calls mean three strangers gave you 20 minutes, which is where you hear the objections in their own words. One payment is the only number that cannot be faked: someone paid a company they had never heard of for a first version. Together they say the idea is worth a second month. They do not say it is a business.
The pattern of misses is as useful as the total. Many visitors and few requests: the pitch or the price on the page is wrong. Requests but no calls: the wrong audience, or no reason to talk. Calls but no payment: the price, the proof or the product is not there yet. Each has one fix, and you change one thing at a time.
How do you get 300 real visitors in 30 days?
Three channels you can count. Search ads with a daily cap: at USD 5.42 a click, 300 clicks cost about USD 1,600, and at a 6.1% click-through rate the ads need about 4,900 views, which keyword volume decides. Named buyers messaged one by one: a list of 100 people at the 3.43% average yields three or four replies, and a small personal list at 5.8% yields about six; 58% of replies come from the first email, so the first message carries the weight. Your own audience: clients, newsletter, past buyers. Do not count friends.
What can a 30-day test not prove?
Retention. A buyer in week two tells you nothing about month six. Unit economics: 19% of the failed companies in the CB Insights set cited unsustainable unit economics, which only shows when the cost of winning a customer is set against what they pay over time. Seasonality, repeat purchase and whether month two converts like month one are all outside the window. A positive verdict buys the next month of spend, and nothing more.
What do you write down before day 1?
- The price on the page and a checkout that takes money.
- One channel you can count, with a daily spending cap.
- The four numbers and the bar for each, written before the first visitor.
- The decision rule: continue, change one thing, or stop, with the date you decide.
- Who reads the numbers every morning.
Sources
- cbinsights.com/research/report/startup-failure-reasons-top/
- localiq.com/blog/search-advertising-benchmarks/
- unbounce.com/conversion-benchmark-report/
- unbounce.com/conversion-benchmark-report/professional-services-conversion-rate/
- instantly.ai/cold-email-benchmark-report-2026
- woodpecker.co/blog/cold-email-statistics/
- cofounderdesk.com/
Cofounder Desk builds your product as real software, puts it live on your own domain and sells it in your name, and its pay comes out of your sales: from USD 2,000 once for the build, then a share of what we sell. If you do not sell, we do not get paid.