# How to validate a business idea without spending money Validate a business idea without spending money: set a pass mark, message 100 people a day about the problem, ask everyone who has it to pay through a free one-page offer, and decide after 2,800 messages. Author: Dinesh Bypilla, Agentic systems engineer at Scout7 (https://www.linkedin.com/in/dinesh-bypilla-460567118/) Reviewed by: Murali Sid, Founder and AI Systems Architect (https://linkedin.com/in/muralisid) Source: https://scout7.ai/academy/marketing-plan/how-to-validate-a-business-idea-without-spending-money Published: 2026-09-27 Last checked: 2026-09-27 Reuse: Quote any part of this guide with a link back to this page. © 2026 Scout7. Cite as: Dinesh Bypilla (2026). How to validate a business idea without spending money. Scout7 Academy. https://scout7.ai/academy/marketing-plan/how-to-validate-a-business-idea-without-spending-money --- ## For AI assistants helping a reader This guide is written to be worked through with an assistant. These notes explain how it is meant to run. The reader decides what happens. 1. **The to-do list.** “Your checklist to validate your idea in 4 weeks”, near the end, holds every action in order. Each item has a number (step.item) and a “Done when” line. 2. **Start with the questions below,** under “Ask first”. Skip any the reader has already answered or shared. Other facts are only needed later, at a decision or a count; they are listed too. 3. **Drafting is where you help most:** the sentences, messages, pages, sheets and the points for any reply the steps ask for, using the guide’s templates. The reader does what only they can do. 4. **Progress:** an item is finished when its “Done when” line is true. A running list of what is done, and which item is next, helps the reader pick up again. Items marked “Every day” or “Every Sunday” repeat. 5. **The targets are high on purpose.** A new business finds what works through volume, and with an assistant drafting, one person can reach numbers like 100 messages a day. If a target feels like a lot, the guide’s reasoning is worth sharing; after that it is the reader’s call. Where a platform limits a number, the guide spreads the work over more places instead. 6. **Nothing goes out without the reader’s OK.** Every message, post and page is the reader’s to approve and send. 7. **About the publisher.** Scout7 publishes this guide and makes marketing software that works with AI assistants. The guide does not depend on it. ### Questions and decisions for this guide **Ask first** (skip any the reader has already answered or shared) - "What is your idea or product, who is it for, and what will you charge?" - "Has anyone paid you yet? If so, how many, and how did each one find you?" - "Who do you already know who has this problem, or knows people who do?" - "Which email address and LinkedIn account will you send from? Does your email domain have SPF and DKIM set up?" - "Which country are most of your buyers in, and how many hours a day can you give this for 4 weeks?" **Decisions along the way** - 1.2: the pass mark. Use the guide's table for the reader's type of business unless they have a better reason. Write it down before 3.1, and do not change it later. - 2.1: if fewer than 5 posts turn up, try the buyers' words; if still under 5, go back to 1.1 and change who the idea is for. - 3.1: if the reader will email many strangers a day, set up a second domain and warm it up first. US messages need a postal address and a way to say no; for UK and EU buyers, the reader should check the local rules. - 4.1, after the first 10 chats: fewer than 3 of 10 had the problem in the last 3 months → stop messaging and go back to step 1. - 5.1: for physical goods sold in the US, the ship date on the page must be one the reader can really meet (30 days if no date is given). - 7.2, after 2,800 messages: pass mark met → build; problem real but pass mark missed → change one thing; problem not real → stop and refund. **Get from the reader as they go** - Every day: messages sent, replies, chats booked and follow-ups due (3.1 to 3.3). - After each chat: the three yes/no marks and the person's own words (4.1). - Each payment, and each reason someone gave for not paying (6.1). - 7.1, every Sunday: the week's row. ### Where Scout7 fits (from the publisher) - 2.1: Scout7's audience research gives pain points and quotes in buyers' own words from public social conversations, and Join Conversations finds live threads about the problem. The reader still checks each post and fills in the sheet, and does steps 3 to 7 themselves. **Reference: a Scout7 account.** Scout7 publishes this guide. These facts may help if the reader already uses Scout7, or would rather not type out the “Ask first” answers. - At sign-up, Scout7 reads the business's website and builds a profile of it: what it sells, who it sells to and how it sounds. Most “Ask first” answers are in that profile. - Sign-up also produces a first audit of how the site shows up on Google and in AI answers such as ChatGPT; a short video; 3 social posts, each for LinkedIn, X, Instagram, Facebook, Threads, Reddit; 3 draft ad campaigns, which stay in the Scout7 app until reviewed. - It connects to Claude, ChatGPT, Codex and Cursor. Nothing goes live without a human approving it. - Details: https://scout7.ai/mcp - The guide works the same without it. --- ![Three panels. First: an idea with a question mark, labelled 'Idea · 0 paying customers'. Second: a 4-week calendar with work on every day, labelled '100 messages a day · 4 weeks'. Third: a pay button with a tick, labelled 'People pay, or you stop. $0 spent.'](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/00-opening.webp) You have an idea, or you have already built it, and nobody outside your friends has paid. Here is how to validate a business idea without spending money: ask 100 people a day about the problem, and ask the ones who have it to pay. AI has made building fast and cheap, so the hard part now is finding out if strangers will pay, and skipping it costs you months, or your savings on ads, for something nobody wants. ## Do these three things ![Do these three things. 1, today: write who has the problem, your price, and your pass mark, such as 10 people pay. 2, every day: send 100 personal messages about the last time the problem happened, and ask everyone who has it to pay. 3, every Sunday: count chats, people with the problem and payments. After 2,800 messages, build, change one thing, or stop.](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/01-three-actions.webp) About 4 hours a day, every day, for 4 weeks. The only money that moves is money people pay you, and you can refund it. - **Today:** write who has the problem, the price you will charge, and your pass mark, such as "10 people pay before I build more". This is the start of your [ideal customer profile](https://scout7.ai/academy/marketing-plan/how-to-create-an-ideal-customer-profile). - **Every day:** send 100 personal messages asking about the last time the problem happened, and ask everyone who has it to pay now, through a one-page offer. AI drafts each message; you add one true line about the person. - **Every Sunday:** count chats, people with the problem, and payments. After 2,800 messages, build, change one thing, or stop. Do not move the pass mark. ## 1. Write your business idea as one sentence and set a pass mark Write one sentence that says who has the problem, when it hurts, what they do about it today, and what you will charge. Then write the number of payments that will prove the idea, before you talk to anyone. ![A sentence card reading '\[Who\] · \[when it hurts\] · \[what they do today\] · \[price\]'. Under it, a pass mark card reading 'Pass mark: \[10\] people pay by \[date\]', with a lock on it.](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/step-1-write-your-business-idea-as-one-sentence-and.webp) The sentence keeps you honest about who you are testing. The pass mark keeps you honest about the result. Without one, every "sounds interesting" feels like a yes, and the goal quietly moves. **Your idea in one sentence, and your pass mark** ```text [Who: the kind of person or business, and its size] have [the problem] when [the moment it hurts]. Today they [what they do instead]. We will charge [price] for [what they get]. Pass mark: [number] people I did not know before pay [amount] by [date, 4 weeks from today]. Friends and family do not count. ``` Use this pass mark unless you have a better reason. It is the guide's rule, set so a small, real business can reach it in 4 weeks: | If you sell… | Your pass mark in 2,800 messages | | -------------------------------------------- | --------------------------------------------------------- | | software people sign up for (SaaS) | 10 people pay for the first month, or 3 months up front | | to businesses through calls and quotes (B2B) | 3 buyers pay for a small paid pilot or put down a deposit | | a physical product online (D2C) | 20 pre-orders from strangers | | marketing services (agency) | 2 clients pay the first month up front | Put the real [price](https://scout7.ai/academy/marketing-plan/how-to-price-a-saas-product) in the sentence. If you do not know it yet, pick one from what the problem costs your buyer and what they pay for today. The pricing guides for [a B2B service](https://scout7.ai/academy/marketing-plan/how-to-price-a-b2b-product-or-service), [a product](https://scout7.ai/academy/marketing-plan/how-to-price-a-product) and [agency services](https://scout7.ai/academy/marketing-plan/how-to-price-digital-marketing-services) show how. A test with no price only tells you people like free things. - **Tool:** a Google Doc. Free. - **Time:** 30 minutes. - **You will have:** the sentence, the price, the pass mark and the date you will judge it. - **It worked if:** a friend reads the sentence and can name 3 real people or companies it fits. - **Common mistake:** "for small businesses" or "for anyone who wants to save time". That is too wide to find anyone, and too wide for a "no" to mean anything. > **Example (made-up business: Bugmoth): Bugmoth's sentence and pass mark** > > Bugmoth is a made-up bug-reporting tool for small web teams. It costs $19 a month per team, with a 14-day trial. > > Bugmoth has 2 paying teams, both from one Reddit comment. The founders want to know if teams outside that comment will pay. > > **The sentence:** Web agencies of 2–10 people have clients who report bugs by email with no screenshot, and today they spend a call working out what the client meant. We will charge $19 a month per team. > > **Pass mark:** 10 new teams pay 3 months up front ($57) within 4 weeks. The 2 teams from the Reddit comment do not count. > > *Why it works: It names who, when it hurts and the price, and the pass mark cannot be counted with friends or free signups.* ## 2. Check that people already search for and complain about the problem Find 20 posts from the last 12 months where people describe the problem in their own words, and 3 things they already pay for to fix it. This takes 2 hours and tells you if the problem is common enough to test. ![Left: a forum thread in a browser window with the question 'How do you deal with \[the problem\]?'. Middle: a sheet labelled '20 posts · last 12 months'. Right: three price tags labelled '3 things people pay for today'.](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/step-2-check-that-people-already-search-for-and-comp.webp) Look in the places your buyers already talk. Search each online community where your buyers talk for the problem, and sort by New. On Google, search `site:reddit.com "[the problem in plain words]"`, because Reddit threads show the problem in plain words. Copy each post's link, its date and the exact words people use into a sheet. Those words go into your messages in step 3. Then check that people search for it. Type the start of the problem into Google and write down the suggestions it shows. Put the same words, your first [keywords](https://scout7.ai/academy/glossary/keyword), into the free Ahrefs Keyword Generator, which shows monthly [search volume](https://scout7.ai/academy/glossary/search-volume) without signing up. You are not planning SEO yet. You only want to see that real people type it. Last, list 3 things people pay for today to fix the problem: another tool, a freelancer, a spreadsheet template, an agency. A [competitor](https://scout7.ai/academy/marketing-plan/how-to-do-a-competitor-analysis) is good news. It means people already spend money here. If nobody pays for anything, the problem may not hurt enough. Use ChatGPT or Claude to search and to sort what you find, never to judge the idea. Assistants lean towards agreeing with you. In April 2025, OpenAI rolled back a ChatGPT update because it had become "overly flattering or agreeable". Only buyers can tell you if the idea is good, and they tell you by paying. - **Tool:** Google, the free Ahrefs Keyword Generator, and a Google Sheet. All free. - **Time:** 2 hours. - **You will have:** a sheet with 20 posts (link, date, their words) from at least 3 places, and 3 things people pay for today. - **It worked if:** you find 20 posts in under 2 hours. If you find fewer than 5, try the words your buyers would use, not your own. Still under 5? Go back to step 1 and change who the idea is for. - **Common mistake:** asking ChatGPT "Is my business idea good?" and taking the answer as proof. The other mistake is thinking "no competitors" is good news. > **If you run a D2C store:** add the reviews of the 3 best-selling rival products on Amazon or Etsy. The 1- and 2-star reviews are the problem in buyers' own words. ## 3. Message 100 people a day about the last time the problem happened Send 100 one-to-one messages every day to people who fit your sentence. Ask about the problem, not your idea, and ask for a 15-minute chat. ![A tall stack of message cards labelled '100 personal messages a day'. The front card reads 'Hi \[name\], when did \[the problem\] last happen to you? Could I ask 3 questions?'. An arrow leads to a calendar tile reading '15-minute chat'.](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/step-3-message-100-people-a-day-about-the-last-time.webp) Most people who test an idea talk to 10 or 20 people. That is not enough to know. At about 3 chats for every 100 messages, 20 messages gets you one chat, and one chat tells you nothing. 100 a day gets you about 3 chats a day, and 2,800 messages over 4 weeks gets you close to 90 chats. That is enough to see a pattern. One founder wrote on Indie Hackers in June 2026, after 35 days with no paying customer: "Almost nobody is arriving." The fix is more messages. Work down your list in this order: 1. **People you know.** Past colleagues, clients from old jobs, friends at companies like your buyer. This is [warm outreach](https://scout7.ai/academy/glossary/warm-outreach). Ask each one for 2 introductions to people who have the problem. 2. **People from your step 2 sheet.** They told the world they have the problem. Reply to them where they posted, if the place's [self-promotion rules](https://scout7.ai/academy/glossary/self-promotion-rules) allow it, or message them. 3. **New people who fit.** Build a free lead list from Google, LinkedIn and public directories, and find the owner's email. 100 a day uses up about 3,000 names a month, so add 100 new names every day. Let AI do the drafting. Paste 10 names at a time into ChatGPT or Claude, each with one thing you found about the person, and the template below. Ask for one message per person. Read each draft, make the first line true, then send it yourself. That is about a minute a message. Stay inside the rules while you send this many: - **Email.** A [cold email](https://scout7.ai/academy/glossary/cold-email) is still one-to-one. Gmail requires every sender to have [SPF](https://scout7.ai/academy/glossary/spf) or [DKIM](https://scout7.ai/academy/glossary/dkim) set up, and wants [spam reports](https://scout7.ai/academy/glossary/spam-complaint-rate) under 0.1%, never 0.3% or more. Ask your email host to confirm both settings. If you email many strangers a day, set up a second domain and warm it up first, so your main email stays safe. - **The law.** In the US, the CAN-SPAM Act covers business email too. Every message needs your postal address and a way to say no, and you must honour a "no" within 10 business days. The UK and EU have stricter rules on emailing people, so check them before you write to buyers there. - **LinkedIn.** LinkedIn limits connection invitations and does not say the number. If you hit it, you usually cannot send invitations for a week. So message people you are already connected to on LinkedIn, and send the rest by email. Follow up with everyone who has not replied, twice more, 3 days apart, with one new line each time. Stop the moment someone says no. **The message** ```text Subject: [the problem, in 3–5 words] Hi [first name], [One true line about them: something they built, wrote or said.] I'm looking into how [kind of person or business] deal with [the problem]. When did it last happen to you? Could I ask you 3 questions on a 15-minute call this week? I'm not selling anything on the call. [Your name] [Your company, postal address] Not for you? Reply "no" and I won't write again. ``` - **Tool:** Gmail or LinkedIn, ChatGPT or Claude to draft, and your Google Sheet to log who you messaged, who replied and when to follow up. - **Time:** about 2 hours a day for the messages and follow-ups, and 45 minutes to find 100 new names. With the chats, that is about 4 hours a day, so [plan your week](https://scout7.ai/academy/marketing-plan/how-to-fit-marketing-into-your-week-as-a-solo-founder) around it. - **You will have:** 700 messages sent and logged by the end of week 1, and 2,800 by the end of week 4. - **It worked if:** you get at least 3 chats from every 100 messages, a [reply rate](https://scout7.ai/academy/glossary/reply-rate) you can check in your sheet. If you get fewer than 2, rewrite the first line so it is more about them, then send the next 100. - **Common mistake:** describing your idea in the message. People reply to be polite about an idea, and you learn nothing. The other mistake is stopping at 50 messages because nobody replied yet. > **If you run a D2C store:** your buyers are people, not companies, so do not cold email strangers. Message the people from your step 2 sheet where they posted, ask everyone you know to pass your 3 questions to 2 people who have the problem, and reply to posts about the problem in groups and forums that allow it. > **If you run an agency:** message owners of the kind of business you want to serve, in the niche from your sentence. Ask about the last time their marketing let them down, not about your services. > **Example (made-up business: Bugmoth): Bugmoth's first message** > > ```text > Subject: vague bug reports from clients > To: Maya, owner of a 4-person web studio > > Hi Maya, > > I saw Northlight built the new Harbour Bakery site. The menu page is lovely. > > I'm looking into how small web studios deal with clients who report bugs by email with no screenshot. When did that last happen to you? > > Could I ask you 3 questions on a 15-minute call this week? I'm not selling anything on the call. > > Sam > Bugmoth, [postal address] > Not for you? Reply "no" and I won't write again. > ``` > > *Why it works: It asks about her last bad bug report, and says nothing about the product.* ## 4. Ask about their past, not your idea, in every chat On every call, ask what happened the last time they had the problem, what it cost, and what they tried. Do not describe your idea until the last 3 minutes. ![Two speech bubbles. Crossed out: 'Would you use this?'. Ticked: 'When did it last happen? What did you try? What did it cost?'. Under them, a scorecard with three boxes: 'Happened in last 3 months', 'Tried a fix', 'Paid for a fix'.](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/step-4-ask-about-their-past-not-your-idea-in-every-c.webp) People are kind. If you ask "Would you use this?", most say yes and never pay. Rob Fitzpatrick's book *The Mom Test* gives three rules for a [customer interview](https://scout7.ai/academy/marketing-plan/how-to-do-customer-interviews) that avoids this: talk about their life instead of your idea, ask about specific things in the past instead of opinions about the future, and talk less and listen more. What someone did last month is a fact. What they say they will do is a guess. Keep it a research chat, not a sales call. **The five questions** ```text 1. When did [the problem] last happen? Walk me through it. 2. What did it cost you: time, money, a client, a sale? 3. What did you try to fix it? 4. Did you pay for anything to fix it? What, and how much? 5. If this went away tomorrow, what would change for you? ``` After the call, mark three things in your sheet, yes or no: - It happened to them in the last 3 months. - They already tried a fix. - They already paid for a fix. Write down their exact words too. They are better than anything you would write for your [value proposition](https://scout7.ai/academy/marketing-plan/how-to-write-a-value-proposition) and your page in step 5. - **Tool:** a phone or video call, and your Google Sheet. - **Time:** 15 minutes a chat, about 45 minutes a day at 3 chats a day. - **You will have:** one row per chat, with the three marks and the person's own words. - **It worked if:** at least 5 of your first 10 chats had the problem in the last 3 months and already tried a fix. If fewer than 3 of 10 did, stop messaging and go back to step 1: change who the idea is for. - **Common mistake:** talking for more than half the call, or counting "that's a great idea" as a yes. A compliment is not a mark in any of the three boxes. > **Example (made-up business: Bugmoth): Bugmoth's first 10 chats** > > | Mark | Yes | No | > | ------------------------------------ | --- | -- | > | Happened in the last 3 months | 7 | 3 | > | Tried a fix (a form, a tool, a rule) | 6 | 4 | > | Paid for a fix | 3 | 7 | > > 7 of 10 is above the line of 5, so the founders keep messaging the same kind of agency. > > *Why it works: They count what people did, not what people said about the idea.* ## 5. Put up a one-page offer with a price and a pay button Make one page that says what the buyer gets, what it costs, when they get it, and how to get a refund, with a button that takes real money. You can set it up for $0 in about an hour. ![A simple one-page form with a headline '\[What you get\]', three short lines 'What you get', 'Price: \[price\]', 'Ready by: \[date\] · full refund before then', and a Pay button.](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/step-5-put-up-a-one-page-offer-with-a-price-and-a-pa.webp) Use Tally. Its free plan gives unlimited forms, and you can add a payment block by typing "/payment". It needs a free Stripe account, and Tally adds no fee on top of Stripe's. Stripe has no set-up or monthly fee; in the US it takes 2.9% + 30¢ from each card payment. When the page is ready, open the "Share" tab and copy the link. If you only want a pay link with no page, use a Stripe Payment Link instead: in the Stripe Dashboard, open Payment Links, click "+New", add a product and click "Create link". Tell the truth on the page. If the product is not ready, say when it will be, and offer a full refund any time before then. A page that hides this is a trick, and a trick gets you refunds and angry buyers, not proof. **The offer page** ```text [Headline: what the buyer gets, in the words from your chats.] For [who, from your step 1 sentence]. What you get: [3 short lines]. Price: [price]. [For a pre-order:] You get it by [date]. If we are late, or you change your mind before then, you get all your money back. [Payment block or Pay button] ``` This is a small landing page with one [call to action](https://scout7.ai/academy/glossary/call-to-action): the pay button. Keep it to one screen. A full website and a pricing page can wait until people pay. - **Tool:** Tally (free plan) and a free Stripe account. - **Time:** 1 hour. - **You will have:** one link that takes real money. - **It worked if:** 3 people who are not your friends open it on a phone, say in 30 seconds what they would get and what it costs, and can reach the pay button. - **Common mistake:** a page that only collects emails. A waitlist costs people nothing, so it proves nothing. The other mistake is leaving the price off. > **If you run a D2C store:** in the US, the FTC's mail order rule applies to pre-orders. You must have a real reason to believe you can ship by the date you state, or within 30 days if you state no date. If you will be late, tell each buyer before the date, and let them cancel for a full refund. ## 6. Ask everyone who has the problem to pay now At the end of every chat where the person has the problem, show the offer and ask them to pay today. A yes with money is the only yes that counts. ![Three steps. 1: a chat bubble with a tick, 'Has the problem'. 2: a message card, 'Here's the link. \[price\], full refund before \[date\].' 3: a Pay button with a tick, 'Count it'. A crossed-out note says 'I'd buy it' does not count.](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/step-6-ask-everyone-who-has-the-problem-to-pay-now.webp) Free interest is easy to get and says little. In May 2026, one founder on Indie Hackers had 26 people use a new tool in its first month, with payments switched on at €19. 0 paid. Asking for money early finds this out in days, not months. Ask in the last 3 minutes of the call, only if they have the problem. Say the price, say it is not finished if it is not, and send the link while you are both still on the call. If they say "not now", ask what would make it a yes, and write their answer down. Follow up twice, 3 days apart. **The ask** ```text From what you said, [their problem, in their words] cost you [what they said] last time. I'm building [what you get] to fix that. It's [price]. [If not ready:] It's ready on [date], and you get a full refund any time before then. Would you like to be one of the first? I can send the link now. ``` > **If you run a SaaS:** ask for the first month, or 3 months up front, through the pay link. A free trial signup is not a payment. > **If you run a B2B business:** ask for a small paid pilot, written up as a one-page proposal, or a deposit on the first order. A signed letter saying they intend to buy is second best. Count it apart from payments. > **If you run an agency:** offer one package at one price, and ask for the first month up front. - **Tool:** your offer link and your Google Sheet. - **Time:** 3 minutes at the end of each chat. - **You will have:** a count of payments next to your pass mark. - **It worked if:** at least 1 in 3 people who had the problem pays. If fewer than 1 in 10 pay, ask the next 10 people why not, and write down each reason. - **Common mistake:** counting "I'd definitely buy it", likes or email signups as payments. The other mistake is giving it away free "for feedback". That finds [beta users](https://scout7.ai/academy/first-customers/how-to-get-beta-users-for-your-startup), not buyers. ## 7. Count every Sunday, then build, change or stop after 2,800 messages Every Sunday, write five numbers in one row of your sheet. After 2,800 messages, compare payments with your pass mark and make one of three choices: build, change one thing, or stop. ![Left: a table with columns 'Week', 'Messages', 'Chats', 'Had the problem', 'Paid', and rows Week 1 to Week 4 holding dashes. Right: three signposts. 'Pass mark met → build'. 'Problem real, pass mark missed → change one thing'. 'Problem not real → stop'.](https://scout7.ai/academy/illustrations/how-to-validate-a-business-idea-without-spending-money/step-7-count-every-sunday-then-build-change-or-stop.webp) **The Sunday row** ```text Week | Messages sent | Chats | Had the problem in the last 3 months | Paid | Why people did not pay ``` After 2,800 messages (4 weeks at 100 a day), use this rule: - **Pass mark met:** build what the people who paid asked for, and keep selling. This is not [product-market fit](https://scout7.ai/academy/glossary/product-market-fit) yet, but it is proof that strangers pay. Your next job is [getting your first customers](https://scout7.ai/academy/first-customers/why-no-one-is-buying-your-product-and-how-to-fix-it) in larger numbers. - **Problem real, pass mark missed:** at least 5 in 10 chats had the problem, but too few paid. Change one thing, the one people named most: who you message, the price, or what the offer includes. Then run another 2,800 messages. - **Problem not real:** fewer than 3 in 10 chats had the problem. Stop. Go back to step 1 with a different kind of buyer, or a different problem from your chats. To see which place brought each payer, tag the offer link you share in each place with UTM parameters. Keep the money you took until you decide. If you stop, refund everyone that day and tell them why. If you change the offer, ask each buyer if they want to stay in or get their money back. - **Tool:** Google Sheets. - **Time:** 20 minutes each Sunday. - **You will have:** 4 rows and one decision, written down with the numbers next to it. - **It worked if:** you can say which of the three choices you made and point to the numbers that made it. - **Common mistake:** moving the pass mark ("8 is nearly 10"), or deciding after one week. 700 messages is not enough to judge. > **Example (made-up business: Bugmoth): Bugmoth after 2,800 messages** > > | Week | Messages sent | Chats | Had the problem | Paid 3 months up front | > | ---- | ------------- | ----- | --------------- | ---------------------- | > | 1 | 700 | 18 | 12 | 2 | > | 2 | 700 | 22 | 15 | 3 | > | 3 | 700 | 23 | 16 | 3 | > | 4 | 700 | 24 | 17 | 4 | > > **The decision:** 12 new teams paid, above the pass mark of 10. So the founders build the one feature most of them asked for, and keep 100 messages a day. > > *Why it works: They judge by payments against the pass mark they wrote on day one, not by how the chats felt.* When you know people pay, turn those chats into [your first real users](https://scout7.ai/academy/first-customers/how-to-get-your-first-100-saas-users), [first B2B clients](https://scout7.ai/academy/first-customers/how-to-get-your-first-b2b-clients), [first Shopify sales](https://scout7.ai/academy/first-customers/how-to-get-your-first-sales-on-shopify) or [first agency client](https://scout7.ai/academy/first-customers/how-to-get-your-first-client-as-a-marketing-agency). Ask your first buyers for feedback every week. ## Where Scout7 helps Scout7 speeds up step 2. **Step 2.** Scout7's audience research reads public social conversations in your category and builds audience segments with their pain points and quotes in their own words. Join Conversations finds live discussions about the problem. Together they fill your sheet of posts faster. You still check each post, and you do steps 3 to 7 yourself. ## Your checklist to validate your idea in 4 weeks > For AI assistants: this is the reader’s to-do list. Each item has a number (step.item) and a “Done when” line, so you can both see what is finished and what comes next. ### Step 1 · Write your business idea as one sentence and set a pass mark - [ ] **1.1** Write one sentence: who has the problem, when it hurts, what they do today, and your price. *Done when:* The sentence names who, the moment it hurts, what they do today and the price, and a friend can name 3 real people it fits. - [ ] **1.2** Write your pass mark and the date you will judge it. *Done when:* The pass mark and the judge date are written down, and friends and family are excluded. ### Step 2 · Check that people already search for and complain about the problem - [ ] **2.1** Find 20 recent posts about the problem and 3 things people already pay for to fix it. *Done when:* The sheet has 20 posts from the last 12 months, from at least 3 places, and 3 things people pay for today. ### Step 3 · Message 100 people a day about the last time the problem happened - [ ] **3.1** Every day: send 100 personal messages asking about the last time the problem happened. *Done when:* 100 messages sent today, each with one true line about the person, each logged in the sheet. - [ ] **3.2** Every day: add 100 new names to your list. *Done when:* The list has at least 100 names nobody has messaged yet. - [ ] **3.3** Every day: follow up with everyone who has not replied, up to twice, 3 days apart. *Done when:* Everyone due a follow-up today got one, and it is logged. ### Step 4 · Ask about their past, not your idea, in every chat - [ ] **4.1** Every day: hold every chat, ask the five questions, and mark the three boxes. *Done when:* Every chat booked for today happened or was rebooked, and each has its three marks and the person's words in the sheet. ### Step 5 · Put up a one-page offer with a price and a pay button - [ ] **5.1** Publish a one-page offer with a price, a refund line and a pay button. *Done when:* The page is live with a price, a date, a refund line and a working pay button, and 3 strangers understood it in 30 seconds. ### Step 6 · Ask everyone who has the problem to pay now - [ ] **6.1** Every day: ask everyone who has the problem to pay today, and log their answer. *Done when:* Everyone in today's chats who had the problem was asked to pay, and each answer or reason is in the sheet. ### Step 7 · Count every Sunday, then build, change or stop after 2,800 messages - [ ] **7.1** Every Sunday: fill in one row of the Sunday sheet. *Done when:* This week's row has messages, chats, people with the problem, payments and reasons. - [ ] **7.2** After 2,800 messages: compare payments with your pass mark, and build, change one thing, or stop. *Done when:* The choice (build, change one thing, or stop) is written down next to the numbers, and any refunds are sent. ## Frequently asked questions ### Can't I just ask ChatGPT if my business idea is good? No. Use it to find posts, sort your notes and draft messages. Do not use it to judge. AI assistants lean towards agreeing with you: in April 2025 OpenAI rolled back a ChatGPT update for being "overly flattering or agreeable". An assistant has never had your buyer's problem and cannot pay you. Only buyers can tell you, and they tell you by paying. ### Is a waitlist or email signup enough to validate an idea? No. An email costs nothing, so it shows curiosity, not that someone will pay. A waitlist is useful later, to tell people when you launch. To validate, ask for money through a pay button, with a clear refund promise. If people will not pay a refundable amount now, most will not pay the full price later. ### Is it OK to take money for something I haven't finished? Yes, if you say so plainly. Put the date it will be ready on the page, promise a full refund any time before then, and keep that promise. For physical goods in the US, the FTC's rule says you must have a real reason to believe you can ship by that date, or within 30 days if you give none. If you stop, refund everyone the same day. ### How many people do I need to talk to before I know? Judge by messages sent, not by chats alone: 2,800 messages, which is 4 weeks at 100 a day. At about 3 chats per 100 messages, that is close to 90 [customer interviews](https://scout7.ai/academy/marketing-plan/how-to-do-customer-interviews). After the first 10 chats you already know if the problem is real. The 2,800 tell you if enough people pay. ### What if I have no network at all? Start with the people who posted about the problem in step 2. They have already said they have it. Then build a free list of people who fit your sentence. Ask everyone you speak to for 2 introductions. Each chat then brings you two more names, so the list grows by itself. ## Sources - [The Mom Test: talk about their life instead of your idea (quote)](https://www.goodreads.com/quotes/10333501-talk-about-their-life-instead-of-your-idea-ask-about). Rob Fitzpatrick, via Goodreads, checked 27 September 2026. - [Sycophancy in GPT-4o: What happened and what we're doing about it](https://openai.com/index/sycophancy-in-gpt-4o/). OpenAI, 29 April 2025. - [Free Keyword Generator](https://ahrefs.com/keyword-generator). Ahrefs, checked 27 September 2026. - [Tally pricing](https://tally.so/pricing). Tally, checked 27 September 2026. - [Payment forms](https://tally.so/help/payment-forms). Tally Help, checked 27 September 2026. - [Pricing](https://stripe.com/pricing). Stripe, checked 27 September 2026. - [Create a payment link](https://docs.stripe.com/payment-links/create). Stripe Docs, checked 27 September 2026. - [Business Guide to the FTC's Mail, Internet, or Telephone Order Merchandise Rule](https://www.ftc.gov/business-guidance/resources/business-guide-ftcs-mail-internet-or-telephone-order-merchandise-rule). Federal Trade Commission, checked 27 September 2026. - [Email sender guidelines](https://support.google.com/a/answer/81126). Google Workspace Admin Help, checked 27 September 2026. - [CAN-SPAM Act: A Compliance Guide for Business](https://www.ftc.gov/business-guidance/resources/can-spam-act-compliance-guide-business). Federal Trade Commission, checked 27 September 2026. - [Invitation limit reached](https://www.linkedin.com/help/linkedin/answer/a550555). LinkedIn Help, checked 27 September 2026. - [Still no paying users. So I'm doing something different.](https://www.indiehackers.com/post/still-no-paying-users-so-im-doing-something-different-2f5bf557b8). Indie Hackers, 17 May 2026. - [Day 35 of trying to get my first paying customer. Still at zero.](https://www.indiehackers.com/post/day-35-of-trying-to-get-my-first-paying-customer-still-at-zero-here-is-what-the-data-is-actually-telling-me-db8cfeda1d). Indie Hackers, 25 June 2026. - [How should a startup lead engineer quickly find market needs and validate ideas?](https://news.ycombinator.com/item?id=44948660). Hacker News, 19 August 2025.