LaunchList
Start free
For testing an idea

Validate your idea with a waitlist before you build

With LaunchList you test an idea before you build it: one page, a signup form and a count of who joined from which link. A signup shows interest, not payment, so the next test asks for money.

Start your testHow idea validation works

Free for your first 100 contactsNo credit card needed

At a glance

Updated

What it is
A demand test for a product that does not exist yet: one honest page with the promise, a form, and numbers on who joined and where they came from.
Starting page
The starting page is Mesa, made for reservations: a large product picture, the details, three price cards and a count of reservations that only ever shows real signups.
What a signup proves
A signup proves interest that cost the person nothing. Whether they will pay is a second test, and LaunchList takes no payments itself.
Two versions
Each page shows one version to everyone, with no traffic split. A second promise runs after the first, or as a separate project: 1 project on the Free plan, 3 projects on the Collect plan.
Sources
The campaign tags on a shared link are saved with each signup it brings, on every plan. Ranking channels, landing pages and referring sites in one report takes the Grow plan.
Price
No charge for your first 100 contacts, with 28 days of analytics history on the Free plan; a bigger plan looks further back.
How it works

How founders test demand before they build.

Each way asks people for a little more than the one before, so each answer is harder to give out of politeness.

  • Problem interviews

    Ask people who have the problem about the last time it happened, what they tried and what they disliked about it. Eric Migicovsky's talk for YC names three usual mistakes: talking about your idea, talking about hypotheticals and talking too much.

    Source: Y Combinator: Eric Migicovsky on talking to users, September 4, 2019

    Use it when

    You cannot yet say who has the problem, or in what words, so you do not know what a page should promise.

  • A page for the unbuilt product

    One page makes the promise and asks for an email, and strangers answer with what they do instead of what they say. It costs them almost nothing, so it shows interest and no more.

    Use it when

    You know who has the problem and can send a few hundred of them, people who have never heard of you, to one link.

  • A pre-sale through a payment link

    Stripe Payment Links make a hosted checkout page without code, for a product or a subscription, and can cap how many are sold. Refunds need no code either.

    Source: Stripe Docs: Payment Links

    Use it when

    People have signed up at the price your page showed, and you want to know how many of them will actually pay it.

Where LaunchList fits

Talk to a handful of people first, so the page promises what they told you in their own words. Send strangers to the page next, and add the payment link after the signup only once the page passes.

When to use it

Is a landing page the right way to test your business idea?

A good fit when

  • You can name who has the problem and where they spend time online.
  • You can reach a few hundred of those people who do not know you, through communities, newsletters or a small ad budget.
  • You want to compare two promises or two audiences, one after the other.
  • You want the people who say yes to become your first users if you build.

Not the right fit when

  • Your buyers are a few dozen companies you can name: call them and ask for a signed letter of intent.
  • The open question is whether people can use the product, or keep using it: that takes a prototype or a working build.
  • Implying the product exists could harm someone, as with health, money or legal advice: test the problem with interviews.
Set it up

An idea validation landing page in three steps.

Before the first step, write down three numbers: your pass mark (the share of visitors who must sign up for the idea to be worth building), the share at which you would drop it, and how many visits you will wait for.

  1. Write the promise on one page

    Start from Mesa and write the outcome you promise as the headline. For the first run, delete the Pricing section ("Add a block here" brings one back), so a low result has one cause: the promise.On every plan

  2. Post tagged links where strangers gather

    Add a source tag to the end of each link you post, such as utm_source=reddit on the one you share on Reddit. In Signups, filter by "UTM source" to count each place on its own, and read "Page visits that signed up" in Analytics against your pass mark.On every plan

  3. Add the price, then the payment link

    Once the page clears your pass mark, add a Pricing section with your real price and a "Single choice" question asking which plan each person would pick. Then set the form's "After submission" to "Custom URL" with a Stripe or Gumroad payment link as the "Redirect URL". Read your referral numbers first: new signups now skip their status page.On every plan

In the docs: building your signup form, what people see after they sign up and reading your analytics.

Real examples

Two ideas strangers signed up for before they were finished.

  • Buffer

    2010, before the first build

    Source:Joel Gascoigne on the Buffer blog, February 16, 2011

    What they did

    Joel Gascoigne tweeted a two-page site for a tool to queue up tweets and asked what people thought; a few left their email. He then added a page showing pricing between the two, and people still clicked through and gave their email, while a small number clicked a plan that cost money. He built the first version over seven weeks of evenings and weekends, and his first paying customer came within four days of launch.

    Take from it

    Test interest first, then put the price in the way and see who still signs up. A question asking which plan they would choose records the answer with each signup.

  • Dropbox

    Private beta, as told in 2011

    Source:TechCrunch, from The Lean Startup, October 19, 2011

    What they did

    The software could not be shown working as a prototype, so Drew Houston made a three-minute video of how it was meant to work, narrated over his screen, with about a dozen in-jokes for Digg's audience. It sent hundreds of thousands of people to the site, and the beta waiting list went from 5,000 to 75,000 people overnight.

    Take from it

    When the product is too hard to build just to find out, show how it would work in a video or mockups and let the signups answer. Aim the first showing at one community, in its own language.

The details

What keeps the result honest, and what comes after it.

  • Your own tries left out

    Test mode keeps your trial signups out of signup totals. While it is on, every signup is marked as a test, and it turns itself off after 24 hours. Visits to the published page still count in Analytics.On every plan

  • Bots kept out of the count

    Each signup is scored as it arrives, and the doubtful ones wait in review, outside every number, until you let them in or block them.On every plan

  • Confirmed signups counted apart

    A waitlist form asks each person to confirm their email, and Analytics shows the share who did, a stricter reading than the raw count.On every plan

  • Referrals as a second signal

    Each status page carries a link to share, and Analytics counts the signups that came through one. Someone who brings a friend wants the product more than someone who only joined.On every plan

  • Where signups come from

    Channels, landing pages and referring sites, ranked for the range you pick, so each source shows its share of the signups at a glance.From the Grow plan

  • Tell the list what you decided

    Email everyone who signed up whether you are building it, and when, within your plan's emails a month.From the Collect plan
FAQ

Idea validation: visits, prices and proof.

How many visitors does a demand test need?
Enough that the rate stops moving. On 100 visits, a page where 10% signed up could really sit anywhere from about 6% to 17%; on 400 visits the range narrows to about 7% to 13% (a 95% Wilson interval). So choose a visit count before you start, a few hundred for most tests, and count only visitors who do not know you.
How long should I run the test before deciding?
Until it reaches the number of visits you wrote down at the start, not until the first good day. If weeks pass and you cannot get there, that is a result too: you have no way to reach the people with the problem yet. On LaunchList's Free plan, Analytics looks back 28 days, and every signup stays in Signups however long the test runs.
What should a page for an unbuilt product say?
The outcome in the headline, who it is for, three short lines on how it would work, the price if price is part of the test, and one sentence saying it is being built and that joining means hearing when it opens. Leave out quotes, logos and numbers you do not have, because invented proof turns the result into a test of the proof, not the product.
Should I show a price on the page?
Yes, once a first run has shown interest. Asking for an email and showing a price on the very first run leaves a low result with two possible causes. On the second run, show the price and ask which plan each person would choose, then send them on to a payment link to see who really pays.
Is a signup proof that people will pay?
No. A signup shows that someone wanted to hear more, and it cost them nothing. Paying is a separate test: LaunchList does not take payments, but a form can send each new signup on to a Stripe or Gumroad payment page of your own, and the share who pay there is the stronger number.
Can I test two versions of my idea at the same time?
Not on one page: LaunchList does not split traffic between two versions of a page. Run one promise after the other on the same source, or give each its own project and send half your traffic to each link. The Free plan holds 1 project and the Collect plan 3 projects.
How do I know which post or ad brought each signup?
Add UTM tags to the end of every link you share, after a question mark: utm_source=reddit names the place and utm_campaign=round1 names the run. LaunchList keeps the tags with each signup on every plan, and the Signups filter reads UTM source and UTM campaign, so each place gets its own count. A ranked report of channels, landing pages and referring sites is on the Grow plan.
What do I do if the test fails?
First check that enough strangers saw the page. Then change one thing, the promise or the audience, and run it once more on fresh traffic. If it fails again, let the idea go and tell the people who did sign up: email them from LaunchList on the Collect, Launch and Grow plans, or export the list as a CSV file on every plan. Archiving the project stops new signups and frees its place in your plan, and the list stays exportable.

Find out who wants it before you build it.

Write down your pass mark, put one honest page in front of strangers, and let the signups and the payment link decide.

Build your waitlist

Free for your first 100 contactsNo credit card needed