Fake Door Test: How to Run One in 6 Steps, With a Worked Example
A fake door test shows real people a button, menu item, ad or flyer for a product you haven't built, then counts who acts on it. Every click lands on an honest note: it isn't ready yet, here is why you asked, leave your email to hear first. You set the number that counts as a yes before anyone sees the door.
What a fake door test is
A fake door test advertises a product or feature before it exists, to measure demand. Optimizely’s glossary defines the painted door test this way and gives “fake door test” as another name for it1. Fake door is also called painted or trap door.
A person sees the offer, clicks it, reads a reveal that says it isn’t built yet, and decides whether to join a list. A prototype is another method, as are concierge and paid pre-orders.
How it works: the door, the click, the reveal
Inside a product, the door can be a button, menu item, toggle, banner or pricing tier for something not built yet. Outside it, the door is a landing page behind an ad, an email link, or, for a local service, a door hanger with a QR code.
| stage | the person | you count |
|---|---|---|
| Door | sees the offer | impressions |
| Click | acts on it | clicks |
| Reveal | learns it isn’t built | answers to one question |
| List | leaves an email or picks a plan | commitments |
In 2010, Buffer’s founder Joel Gascoigne put up a landing page that led to a plans-and-pricing page and then an email sign-up, and started building only after people signed up and some clicked paid plans2. A click on a paid plan says more than an email.
When to run one, and when not to
A door measures what people do with an offer, not what they guess they might do. It takes days, because you draw a door instead of building a product. Its cost is trust: each click meets a small letdown, so keep the window short and show the door once per person.
Skip the door when:
- The feature is quicker to build than to fake. Build it and measure use.
- The flow is critical or regulated (checkout, health, money). A dead end there costs too much.
- You need to know whether people can use it. Make a clickable prototype.
- You need the market’s size. A door counts only who you reached; see market analysis.
- You need the reason behind a choice. Use customer interview questions.
How to run a fake door test in 6 steps
1. Pick the riskiest assumption
Rate each risk of your idea from 1 to 3 on likelihood and on impact, the scale our idea check uses, and multiply the two. Test the highest-scoring risk a door can answer: demand, willingness to pay at a price, or a choice between two offers. “Can we deliver” goes to another test. Write it on your lean canvas, then as one sentence:
“If we show [door] to [segment] for [window], at least [pass line] will [commit], at a cost of no more than [ceiling].”
Set a fail line below the pass line. A result between the two gets one rerun with one change. On a team, the product owner sets both lines and decides; marketing places the door, design draws it, an engineer logs events, and support answers the list.
2. Pick the door and who sees it
Use one channel and one segment, defined by behavior, from your ideal customer profile. An ad door also builds a list of the people it found. Keep a control group that never sees the door. Size the reach: people to reach = commitments the pass line needs ÷ (expected click-through rate × commitment rate). Both rates are your assumptions.
3. Write copy that invites
State the offer from your value proposition: one benefit, one price, one action, and no claim that the product is live. “Get early access” names nothing; “Reserve a spot on the first weekly route” names the job. Freeze the copy for the window, since an edit splits the data into two smaller tests.
4. Build the reveal
On the first click, the reveal says what is coming and why you asked, says what an email gets them (first notice, nothing to pay), asks one question such as “What made you click?”, and offers an optional email field. Give no launch date. A date promises a product that may never ship.
5. Track and run
Log who clicked, when, where and how often. Use analytics events in product, a form for the list, UTM tags on outside links, and one QR code per flyer batch. Run the full window.
6. Decide and follow up
Compare the counts with your lines. Within a week of closing, tell the list what you decided and take the door down.
Set the pass line from your own numbers
No click-through rate is good in general. Set the line from your own economics before launch, using two rates:
- Click-through rate = clicks ÷ people who saw the door.
- Commitment rate = commitments (an email, a picked plan) ÷ clicks.
Read each segment before the total, because one strong segment can hide inside a weak average.
For ads, flyers and landing pages, cost per commitment = test spend ÷ commitments. The door passes at or below the customer acquisition cost (CAC) your model assumes. Only part of those who commit will pay, so the CAC is a ceiling.
For an in-product door, users needed = the feature’s build cost ÷ one user’s contribution over a period you pick. Both numbers are your assumptions. The door passes if its commitment rate, applied to everyone who would see the feature, reaches that count.
Worked example: a door hanger for a dog waste service
This is an example by our method. No door was run, so it sets up the test and reports no result.
The idea in the example: “Dog poop scooping service in suburban Dallas–Fort Worth subdivisions: dog-owning homeowners paying $80–$120 monthly weekly-visit plans, launched via door hangers and neighborhood apps”ex. Our October 15, 2026 check rated it “Worth building”ex with a score of 69ex in the example; 17 searchesex found its 26 sourcesex. In the example it found 9 direct competitorsex. POOP 911 holds 21%ex of visibility, Doggy Doggz 10%ex.
Two top risks are what a door measures. The incumbent, rated 3 on likelihood and 3 on impact in the example: “Doggy Doggz already offers most of this idea in the same area, so subscribers may choose the incumbent”ex. Price scored 2 and 3 in the example: “willingness to pay at $80-$120 a month is unproven”ex. A third risk shapes the copy: customers complain about reliability, “and no competitor addresses it”ex. That is one of the customer pain points a guarantee can answer.
The door is a door hanger with a QR code plus one neighborhood-app post. The copy uses the check’s top demand phrase, “dog poop scooping”ex, and $100 a month, the middle of the idea’s range in the example.
- Bad: “Your weekly cleanup starts Monday.” It says the service is live.
- Good: “Weekly dog poop scooping, $100 a month. Miss a visit and that week is free. Scan to reserve a spot.”
The reveal: “We’re not running routes in your subdivision yet. We’re checking whether enough neighbors want this before we start. Leave your email and we’ll tell you first. Nothing to pay.” Then one question: “What made you scan?”
The pass line, example by our method. Assumption: 400 door hangers in two subdivisions, $200 to print and deliver. The check puts CAC at $30ex in the example. $200 ÷ $30 = 6.7, so the door passes at 7 or more people who scan, pick the $100 plan and leave an email, under 2% of 400 households.
| result | decision |
|---|---|
| 7 or more | pass: run a paid pilot route |
| 4 to 6 | rerun once with one change (drop the guarantee, or try a second price) |
| 3 or fewer (assumption) | fail: change the price or offer and rerun once, or drop it |
Hang the cards on the door. Federal law bars putting mailable matter with no postage paid in a mailbox3.
Stay honest: the copy, the reveal, the money
A door stays honest when the copy invites, the reveal comes on the first click, and no money changes hands. Four US rules mark the edges.
- Money. The FTC’s Mail, Internet, or Telephone Order Merchandise Rule makes a seller ship within its stated time (30 days if none is stated) or notify the buyer and offer cancellation with a prompt refund4. Take emails, never payments.
- Claims. Section 5 of the FTC Act prohibits unfair or deceptive acts or practices in or affecting commerce5. Copy saying the product is available today is a claim you can’t back.
- The list. Under CAN-SPAM, each commercial email needs a working opt-out and a valid physical postal address, and you must honor an opt-out within 10 business days6.
- Ad doors. Google Ads’ misrepresentation policy bans ads for offers that are unavailable or hard to find from the destination7. Check your ad network’s policy first.
Fake door test vs. prototype, pre-order and concierge MVP
| method | measures | pick it when |
|---|---|---|
| Fake door | whether people act on an offer | you need demand before building |
| Pre-order page | whether people pay | you can ship in the stated time |
| Clickable prototype | whether people can use it | demand is known, usability isn’t |
| Concierge MVP | whether the result is worth paying for | you can deliver by hand |
| Interview or survey | what people say they would do | you need the reason |
Your next step
Write down your riskiest assumption and its pass line before you draw the door. The free check finds that assumption for you.
FAQ
Who should run a fake door test?
The product owner sets the pass and fail lines and decides; marketing, design, an engineer and support each own one part.
What are the benefits of a fake door test?
You see what people do, in days, without a build. The cost is some trust per click.
Fake door test or prototype: which one do I need?
A door measures demand; a prototype measures usability. Run the door first if demand is the open question.
What is a good click-through rate for a fake door test?
No universal number exists. Set the pass line before launch from your CAC, or from the feature’s build cost for an in-product door.
Is a fake door test ethical, and is it legal?
Yes when the copy invites, the reveal is instant, and you take no money. Never claim the product exists, and follow CAN-SPAM for the list.
Sources
- optimizely.com, 06/18/2026
- linkedin.com, 01/21/2026
- uscode.house.gov, 06/03/2026
- ftc.gov, 02/03/2014
- uscode.house.gov, 10/18/2026
- ftc.gov, 12/29/2025
- support.google.com, 09/19/2026
Check your idea for free
Describe your idea and the check collects demand, competitors, complaints and prices, with sources.
