Demand testing for unbuilt products
Find out if they would pay — before you build
You publish a waitlist page with the expected launch price in plain sight. Everyone who signs up picks one of three levels: just tell me, I want the founder price, or I need this now. What you read at the end is demand, not curiosity.
How it works
- 1
Describe the product and publish the price
One page: what the product does, who it is for, and the price you intend to charge at launch — always labelled as expected, never pretending the thing already exists.
- 2
Put it in front of people with the problem
A community, a small ad, a forum answer, search. The page can carry search traffic because every idea ships with its own content pages, not just a landing.
- 3
Read demand by level of intent
Each signup arrives with an email, a chosen level and an optional line of context. Twenty people who need it now outweigh two hundred who want a launch email — and that is the number the decision rule reads.
What you get
Three levels of intent
Notify me, founder price, I need this now. The form forces the choice, so the list arrives pre-sorted by how much each person meant it.
Signups in your Slack
Every signup lands in a private channel with the page it came from, the language, and the context line. No dashboard to check.
Pages built for search
Each idea carries its own set of specific pages — costs, comparisons, templates — so people with the problem can actually find it.
The price is always labelled
Every page says the product does not exist yet and calls the number an expected launch price. Testing demand does not require misleading anyone.
No database, no trackers
Static pages, no third-party analytics, nothing stored about your visitors beyond the signup itself.
More than one language
An idea can run in English, Portuguese and Spanish when the market justifies it, each with its own written pages.
What the alternatives actually measure
All four of these will get you a page with an email field on it. Only one of them makes the visitor commit to a number.
| Question | Esquyna Waitlist | Carrd + Mailchimp | Typeform | Prefinery |
|---|---|---|---|---|
| Shows a price before the email field | Required, on every page | Only if you write it in yourself | Possible, rarely done | No — position in queue instead |
| Separates a browser from a buyer | Three tiers, mandatory | One list | Whatever you build | One list, ranked by referrals |
| Entry price | $39/mo expected | $19/yr plus a free tier | from $25/mo | from $99/mo |
| Pages a stranger can find from search | Up to 20 per idea, in 3 languages | One | None — it is a form | One |
| What it optimises for | Fewer, heavier signups | Signups | Response rate | List size and referrals |
| Where signups land | One Slack channel, no database | A mailing list | A results table | A dashboard |
Competitor prices are their published entry tiers as of August 2026. Carrd's $19/yr is the Pro Standard plan; the free tier does not allow forms.
Before you put a price on a page
The checklist below is the one this method applies to every idea before it goes live.
- Name the person, not the segment"Freelance designers who invoice from a spreadsheet" is a person. "SMBs" is a rounding error you can hide inside for a year.
- Write down what they do today insteadIf the honest answer is "nothing, and it is fine", the price will not save you. That is a real result and it arrives free.
- Pick the price before you write the copyCopy written after the price justifies it. Copy written before the price tends to invent a number that flatters the copy.
- Check what the incumbent charges, on their pricing page, todayNot from memory and not from a listicle. Prices move, and a comparison table with a stale number is worse than no table.
- Decide the number that means yes, in writing, before launchOurs: fifty signups and fifteen in the top two tiers inside sixty days, or five urgent-tier signups in any fortnight.
- Decide the number that means noThis is the one everybody skips. Without it, every result is encouraging.
- Publish something a stranger would read without the productA calculator, a template, a comparison with real prices. If the only content is the pitch, nobody arrives from search.
Seven items, and the sixth is the one that actually decides whether the experiment was worth running.
What this would cost
Published before the product exists, on the same terms it asks of every idea.
Expected launch price
$39per month
Founder price$19 per month
This product does not exist yet. The number above is the price we expect to launch at, not a commitment. If the waitlist says it is wrong, it changes.
What the price covers
- Unlimited ideas, each with its own price, theme and page set
- Three-tier intent capture on every form
- Content as versioned JSON — a new idea is a data change, not a deploy of new code
- Per-page Open Graph images, sitemaps, hreflang and structured data generated from that data
- Signups delivered to Slack, and nowhere else
What it does not cover
- Email sending. It does not run your launch sequence and is not trying to.
- Analytics. Bring your own, or read the Slack channel.
- Payment collection. Nobody is charged anything before the product exists.
- An opinion about whether your idea is good
The founder price is held for as long as the account stays open. It is the only promise worth making before a product exists.
Three objections, two of which are correct
Two of these are good enough that they should probably stop you. The third is the one I disagree with.
- A price on a page nobody can buy is a fake door, and fake doors are dishonest.
- It depends entirely on what happens after the click. A fake door that says "coming soon!" and quietly harvests the email is dishonest. A page that says the product does not exist, shows the price it would launch at, and tells you what happens next is a survey with a number in it. Every page on this domain says unbuilt above the fold. If that stops being true, the criticism lands.
- I can do this with a landing page builder in an afternoon.
- You can, and for one idea you should. The arithmetic changes at four or five. Fifteen ideas at one page each is fifteen pages that nobody can find; fifteen ideas at ten pages each, in three languages, with reciprocal hreflang and a link graph that has no orphans, is two hundred and something pages — and that is a data problem, not a design problem.
- Waitlist signups do not predict revenue anyway.
- Correct, and the research on this is not flattering. Undifferentiated signups predict revenue badly enough that the correlation is close to useless. What is being claimed here is narrower: that a signup made after seeing a price predicts better than one made without, and that the top tier predicts better than the bottom. That is a testable claim and this site is the test.
Join the waitlist
Same form, same three levels this product asks of every idea it publishes. If demand does not clear the bar, it does not get built.