Fake door test: how to validate an app idea with a reel before you build the MVP

A fake door test shows people the door to a product that does not exist yet and counts who tries to walk through it. It is the cheapest way to learn whether anyone wants your app. In 2026 the cheapest door is a reel: post a demo of the app you have not built, put a waitlist link in your bio, and count who clicks.

This guide is from Hook Studio. Hook Studio is an AI reel maker that turns a short chat into vertical reels for TikTok and Instagram, for founders who want to test an app idea before they build it. Below: what an MVP is, the classic MVP tests and where they come from, and a step-by-step way to run a fake door test with reels.

A fake door test puts up the entrance to something you have not built, such as a button, an ad, a page, or a "link in bio", and measures how many people try to go in. Each click is a person who wanted it enough to act. After the click, you tell them the truth: it is not ready yet, and here is how to get it first.

The idea comes from pretotyping, a set of techniques developed at Google in 2010 by Alberto Savoia and his colleagues. A prototype asks whether you can build something. A pretotype asks whether you should. In his book Pretotype It, Savoia lists the Fake Door as one of the techniques: create a fake entry for a product that does not yet exist in any form.

The rules that keep a fake door honest

• Say it right after the click. The page says the product is being built and offers early access. Nobody should think they bought something.
• Take no money. A click and an email are the signal. Charging for something that does not exist is not a test.
• Follow up. Everyone who joined gets an email when there is news, even if the news is that you dropped the idea.
• Mind who you test on. If your buyers are a handful of companies you need to trust you, talk to them instead.

An MVP, or minimum viable product, is the smallest thing you can put in front of real people to learn whether your idea works. Eric Ries, who popularized the term in The Lean Startup, defines it as "that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort." The term itself was coined in 2001 by Frank Robinson and spread through Steve Blank and Ries.

The word that matters is learning, not product. An MVP does not have to be a small app. Writing about Dropbox in 2011, Ries put it plainly: "In this case, the video was the minimum viable product." A demo video, a page, or a service you run by hand all count, as long as it tests the guess most likely to sink you.

Each classic MVP fakes a different part of the product. Pick the one that tests your riskiest guess, and skip the rest.

People use these names loosely. A fake door usually sits inside something that already exists, like a button in your app or the link in your bio. A smoke test usually means buying or earning traffic to a page and counting who tries to buy. Ries tells of a feature his team spent two weeks building that nobody wanted, and says a simple AdWords smoke test would have shown it much sooner.

Three famous MVPs that were not the product

• Zappos (Wizard of Oz). Nick Swinmurn photographed the shoes in a shoe store, put the photos on his site, and bought the pair from the store at full price when someone ordered. Customers saw an online shoe store. Behind it was one person.
• Buffer (landing page). Joel Gascoigne put up a two-page site to see whether people would even consider the app, then added a pricing page between the two and watched which plan people clicked before leaving an email. His write-up is called "Idea to Paying Customers in 7 Weeks".
• Dropbox (explainer video). Drew Houston posted a short demo of the product, narrated over his screen. In his words, "Our beta waiting list went from 5,000 people to 75,000 people literally overnight."

Dropbox is the one to copy in 2026. Its MVP was a screen demo posted where early adopters already spent their time. Today that place is the For You feed and Reels.

A reel is an explainer video, a fake door, and a smoke test in one, and the feed shows it to strangers for free. The comments tell you what people want in their own words. The link in your bio tells you who wants it enough to leave the app.

Building got cheap. With AI coding tools, people ship a working app in a weekend; we explain how in what vibe coding is. That makes the build the wrong thing to test. If the app takes three days, the slow and risky part is getting strangers to care, and a reel tests that before you spend the weekend.

Prayer Lock shows that the hook, not the code, is the hard part. Its founder, Mau Baron, built the app in about three days with AI. Then he tested formats with up to 40 posts a day across 12 accounts. In his words: "It wasn't until after 40 posts that I finally found a hook that worked."

• The format: a person reacting, with a demo of the app. A competitor was already doing well with it, so he copied the shape, not the content.
• The conversion: "In my case 10K views a day was getting me around two to three hundred downloads a day." He adds that the format is very optimized for downloads, so treat it as his number, not a benchmark.
• The money: he put ad spend only behind organic videos already doing more than 10K views.

Mau built first, and it cost him three days. If your idea takes longer to build, test the hook first: the reel does not need the app to exist, only to look real for a few seconds. The full story is in how Prayer Lock hit $100K a month.

Post a demo of the app as if it were coming soon, send people to a waitlist, and count who goes. This is the order we would use.

1. 1Write the bet in one line. Who has the problem, and what they get. "Students who doomscroll before exams get their apps locked until they finish a study block." If it does not fit in one line, it will not fit in a hook.
2. 2Make something to show. A Figma mock, a clickable prototype, or screenshots of one vibe-coded screen. It has to look like the product for a few seconds. It does not have to work.
3. 3Pick three formats. A person reacting, then the app's screens. A "would you use this?" where someone shows the screen and asks. The problem first, then the demo as the fix.
4. 4Build the door. A waitlist page with one sentence on what the app does, one email field, and an honest line: "We are building this. Join to get it first." Put the link in your bio and end every reel with "link in bio". If your account cannot add a clickable link yet, count comments and DMs instead.
5. 5Tag every source. Give each format its own link, for example by adding ?src=reaction to the end, so your waitlist tool shows which format sent each signup.
6. 6Post 30 to 40 times before you judge the idea. About 10 posts per format, then more of whichever format holds best. One post can be luck either way.
7. 7Answer every comment that asks for the app. Send the link and one of the questions below. These are your first interviews.
8. 8Decide with the rules below. Write your pass mark before you post, not after you see the numbers.

Formats are easier to copy than to invent. Our reel templates post breaks 12 of them down second by second, and viral hooks examples has 50 openings to start from.

Measure what people do, from the cheapest action to the most costly. Views tell you the hook works. Only clicks and signups tell you the idea works.

Count signups per 1,000 views for each format, and compare your formats with each other rather than with someone else's account. We wrote more about why a comment that asks for the app beats a like in "what app is that?".

Change one thing at a time, starting with the cheapest. These are rules of thumb, not laws, and they assume you posted 30 to 40 times.

Write your pass mark before you post, in a number you can count: "If 40 posts bring fewer than N signups, I change the angle." Choose N from what a user is worth to you. When a reel passes it, build the MVP, and keep posting the hook that got you there.

Ask about what people already do, not whether they would use your app. Rob Fitzpatrick's The Mom Testboils it down to three rules: talk about their life instead of your idea, ask about specifics in the past instead of opinions about the future, and talk less and listen more. That is why "would you use this?" makes a good hook and a bad interview question.

1. 1"When was the last time this happened to you?" A specific story means the problem is real. "All the time" with no story usually means it is not urgent.
2. 2"What do you do about it now?" Their current fix is your real competitor, even when it is a notes app or nothing.
3. 3"Have you paid for anything to fix it?" Money already spent is the closest thing to a price test before you charge.
4. 4"What made you tap the link?" The answer tells you which promise to put in the next reel.
5. 5Once they have used the MVP: "How would you feel if you could no longer use it?" This is Sean Ellis's product-market fit question. He benchmarked nearly a hundred startups and found that the ones with traction almost always had more than 40% of users answer "very disappointed".

Hook Studio makes the demo reels, so you can post 40 of them without filming 40 times. You describe the app and the reaction you want. The agent makes a still, then a short clip from the still you like, and adds your prototype's screenshots, the text, and an end card that points to your bio.

These prices leave out the render and the automatic realism checks, a few credits each. Forty hooks are more than a month of the Concept plan (2,300 credits for $20) and fit in Grow (5,800 for $50). Your first $5 costs $3 and gives you 500 credits, enough for three test reels. If you are happy on camera, filming yourself costs nothing.

What it does not do: a real person may still convert better than an AI person, and Mau's winning videos were real people. Use AI reels to find the hook cheaply, then film or pay creators for the winner. Hook Studio does not build your waitlist page or count your signups; that is your own tool. You download the reel, post it yourself, and turn on the AI label, as TikTok and Instagram ask. See how the reels are made on the AI reel maker page.