# The 72-Hour Demand Test **From Chapter 2.3 of *The 4-Hour Side Hustle*.** Seventy-two hours. Under thirty dollars. The page structure, the number that decides, and the obligation that comes with running it live. Surveys measure politeness. This measures intent — because there is exactly one way to find out whether strangers will pay for something, and it's to put a real price in front of them and watch whether anyone reaches for a card. --- ## Before you start You should be here with one or two ideas that scored **above 18** on the Chapter 2.1 scorecard and survived an argument. If you're below that, this test will tell you what the scorecard already told you, and it'll cost you a weekend to hear it again. **Decide one thing now, before you build anything: live mode or test mode.** - **Live** — real charges. Strongest signal, and it comes with a non-negotiable obligation (see below). - **Test** — Stripe test mode. You lose completed-purchase data, but the click rate still counts, and the click rate is the number that matters anyway. If any risk of a real charge would bother you, choose test mode now rather than discovering your discomfort after someone's card is charged at 11pm on a Sunday. --- ## The page — five elements, in this order Forty-five minutes. **Set a timer.** This page is a measuring instrument, not a product launch, and the temptation to polish it will eat your Saturday and teach you nothing. Carrd ($19/year for the pro tier you'll want) or Framer's free tier. ### 1. Headline — the specific outcome Name what the buyer ends up with, not what the product is. "Close your month-end compliance review in one evening" beats "The Compliance Workbook." One line. No subhead unless the headline genuinely can't stand alone. ### 2. Three bullets — what's inside Three. Not five, not seven. Each one concrete enough that a stranger could picture it. Say what it *is*, not what it will do for them — the headline already made that promise, and repeating it reads as though you're padding. ### 3. The price Visible, unqualified, on the page. Not "from," not "starting at," not hidden behind a click. **It must be your actual intended price.** A test at $9 for something you'll sell at $79 measures whether people like cheap things. You already know the answer to that. ### 4. The buy button A Stripe Payment Link. Four minutes to create, no code. One button. If you're tempted to add a second option, a tier, or a "join the waitlist instead" escape hatch — don't. Every alternative you offer is a way for someone to express interest without expressing intent, and interest is the thing you're specifically trying to stop measuring. ### 5. One sentence of credibility Why you're the person selling this. One sentence, first person, no bio, no photo. ### What the page must not have - A hero image - Testimonials you don't have - An FAQ - A countdown timer - A logo you spent an hour on Each of these adds build time and none of them changes the number you're measuring. --- ## Analytics **Plausible** ($9/mo, 30-day trial covers this entire test) or **Fathom**. Not Google Analytics. You need one number, quickly, and GA will bury it under a configuration wizard you'll still be reading on Sunday night. --- ## Traffic — one source, one hundred visitors One relevant subreddit, or a LinkedIn post to your actual network, or a niche Slack or Discord, or a Facebook group. **One source.** Resist adding a second. Two sources give you a muddier read and no more confidence, and you'll spend the analysis wondering which audience produced which behaviour. Drive exactly one hundred visitors. Stop looking at the number until you get there — an early 0-for-14 feels like a verdict and isn't one. --- ## The number that decides Look at the **checkout-click rate**: people who clicked the buy button. Not people who completed a purchase. | Click rate | Verdict | |---|---| | **Above 5%** | Strong. Build it. | | **3–5%** | Viable. Build it, and expect to work on the offer. | | **Under 3%** | Back to 2.1. Something is wrong with the idea, the audience, or the price — and building will not fix it. | Clicks rather than completed purchases, because you may not have a product to deliver yet. A click on a real button at a real price is the closest thing to intent you can measure without shipping. Someone who clicks has decided. **Under 3% is a result, not a failure.** It's four hours that saved you eleven weeks. Record it in your ledger and go back to 2.1 with better information than you had. --- ## If you run it live: the obligation If someone reaches checkout and pays, you owe them something, and it isn't negotiable. **Refund within twenty-four hours, with a personal note.** Tell them the truth: you were testing demand before building, their purchase told you it's worth building, and you'd like to send them a free copy when it's ready. **Set the calendar reminder before you publish the page.** Not after the first sale — before. This is the step that slips, and it slips because you're excited about the sale. This converts into early advocates far more often than you'd expect. People aren't offended by the truth; they're offended by being ignored after paying. Two of Priya's most useful early conversations came from people she refunded in week one, and both bought again at full price. What you must not do is take money for something you don't intend to build, or leave someone paid-up and silent while you decide. That isn't a growth hack. It's theft with better branding. --- ## The part everyone forgets **Save every comment, reply and question the test produces, in a plain document, in the buyer's exact words. Don't paraphrase and don't tidy them up.** You won't need them this week. In Chapter 8 you'll write a sales page, and the single biggest difference between a page that converts and one that doesn't is whether it's written in your vocabulary or theirs. Marcus's buyers said "covenant tracking." He'd been writing "regulatory workflow optimization" for a year. That gap cost him more than the test did. --- ## What a real result looks like Marcus ran it on a workbook at $340, to a commercial-lending subreddit and one LinkedIn post. 118 visitors over four days. Nine clicked through to checkout. Two paid — both refunded within the day with a note — and one replied asking whether there was a version for agricultural loans, which became his second product a year later. A 7.6% click rate at $340 to a cold audience. Worth being precise about what that proved: not that the product was good, since it didn't exist, but that the *problem* was expensive enough for strangers to reach for a card without a phone call. --- ## Checklist - [ ] Idea scored above 18 on the 2.1 scorecard - [ ] Live or test mode decided **before** building - [ ] Carrd or Framer page built in 45 minutes, timer set - [ ] Headline names the outcome, not the product - [ ] Exactly three bullets - [ ] Real intended price, visible and unqualified - [ ] One Stripe Payment Link button, no alternatives or waitlist - [ ] One sentence of credibility, first person - [ ] No hero image, no testimonials, no FAQ, no countdown - [ ] Plausible or Fathom installed — not Google Analytics - [ ] One traffic source chosen, second source resisted - [ ] If live: refund reminder in the calendar **before** publishing - [ ] Exactly 100 visitors driven before reading the result - [ ] Checkout-click rate recorded in the ledger - [ ] Every comment and reply saved verbatim, un-tidied, for Chapter 8