Value Proposition Test - Fake Door

In Brief

A fake door test is a button, link, or advertisement for a product or feature that does not yet exist, placed in front of users to measure interest. You measure how many people click through, producing a click-through rate that reveals real behavioral interest — not just stated preferences. It is commonly used to test new feature ideas within an existing product or to validate a concept with external ads before committing development resources.

Description

Fake door tests are part of the Value Proposition Test family — methods that test demand for a promise by asking participants to commit money, time, data, or actions. The click on a non-existent feature is a deeper-funnel commitment than a top-of-funnel signal because the prospect is already engaged with the product.

A fake door test measures real behavioral interest — clicks, not opinions. You place a button, link, or ad for something that doesn’t exist yet, and count how many people try to use it. The [click-through rate](https://en.wikipedia.org/wiki/Click-through_rate “Clicks divided by views — the percentage of people who saw the door and tried to open it”) tells you whether demand is real. The pattern is one of the canonical pretotyping techniques described by Alberto Savoia in The Right It — a way to test whether you are building the right thing before you spend the engineering budget to build it right.

There are two types:

• Internal fake door — A button or menu item inside your existing product. Tests whether current users want a new feature. Example: adding a “Switch to map view” button on a page that only has a list view.
• External fake door — A paid ad or social post for a product that doesn’t exist yet. Tests whether a target audience is interested in a concept. Example: running a Facebook ad for an online community to see if anyone clicks.

Don’t use fake doors to test essential functionality (like registration flows) or your core value proposition. Use them to test additions and new directions.

Ethics: You’re showing people something that doesn’t exist. Keep it honest. The reveal page should clearly say the feature is in development and offer to notify them when it’s ready. Used in moderation, fake doors save you from building things nobody wants. Overused, they erode trust.

How to

Prep AI Prompt

1. Write your hypothesis.

State what you expect to happen. Be specific: “At least 5% of users who see the button will click it” is testable. “People will be interested” is not.

2. Decide: internal or external.

• Internal (in your product): Best for testing new features with existing users. Requires adding a button or link and a tracking event. Fastest to set up.
• External (paid ads): Best for testing a new concept with a new audience. Requires an ad account and a landing page. More expensive but reaches people who don’t know you yet.
• Hybrid (ad → landing page with fake CTA): Tests both audience interest and messaging. The ad proves they’ll click; the CTA proves they’ll act.

3. Set your success threshold before you launch.

Decide what click-through rate counts as a win. Benchmarks:

• Internal buttons/features: 2–5% CTR is typical for in-product UI elements. Above 5% is strong interest.
• External ads: Industry average CTR for Facebook ads is ~1%. Google Search ads average ~3–5%. Beat the channel average to call it a win.
• Landing page CTAs: 5–10% email signup rate from a cold audience is solid.

If you don’t set a threshold up front, you’ll rationalize any result as success.

4. Set a target sample size.

You need enough views to trust the click-through rate:

• For a rough directional signal: 200+ views minimum.
• For a confident comparison between variants: 500+ views per variant.
• Use a sample size calculator if you need precision.
• If your traffic is very low (under 200 views): Run the test longer, use external ads to supplement traffic, or combine the fake door with qualitative follow-up (interview the people who clicked). A 10% CTR on 50 views is a signal worth investigating, even if it’s not statistically conclusive.

5. Design the fake door.

Rules:

• Make it look exactly like a real feature would look. Same font, same placement, same style. If you make it flashier or bigger than normal UI elements, you’re measuring novelty, not interest.
• Write the label/copy in the same language you’d use for a real feature. No hype.
• If testing multiple variants (different copy, different placement), run them sequentially or split traffic evenly.

6. Design the reveal page.

What users see after they click. Options:

• ”Coming soon” page — Explains the feature is in development. Offers email signup for notification. Best for features you might actually build.
• ”Help us prioritize” page — Short survey asking why they clicked and what they’d expect. Best for early exploration.
• Simple thank-you — “Thanks for your interest. We’re exploring this idea.” Minimal but honest.

Always include: a clear explanation that the feature doesn’t exist yet, and a way for the user to continue what they were doing without frustration.

7. Set up tracking.

• For in-product fake doors: Add a click event to your analytics (Google Analytics, PostHog, Amplitude, Mixpanel). Track both views (impressions) and clicks.
• For external ads: The ad platform tracks CTR automatically. Also track visits to the reveal page and any email signups.
• For both: Record the date range, total views, total clicks, and any segmentation data (user type, device, source).
• Before launching: Test the tracking yourself — click the fake door and confirm the event appears in your analytics. This is the most common reason fake door tests produce no usable data.

Analysis AI Prompt

1. Calculate the click-through rate.

CTR = clicks ÷ views × 100. This is your primary metric.

2. Compare against your pre-set threshold.

• Above threshold: The concept has enough interest to justify further investment. Move to the next validation step (solution interview, pre-sales, or build).
• Below threshold but close: Consider testing a different copy variant, placement, or audience before giving up. A 3% CTR when you needed 5% might mean your messaging is off, not that demand is zero.
• Well below threshold: The concept doesn’t resonate with this audience. Either pivot the concept or try a different audience.

3. Segment if possible.

If you have enough data, break CTR down by user type, device, or traffic source. A feature that gets 8% CTR from power users but 1% from new users tells you something different than a flat 3%.

4. Check for confounds.

• Did something else change in the product during the test period?
• Was there a spike in traffic from an unusual source (press mention, viral post) that diluted the signal?
• Was the fake door placed in a high-traffic area that gets incidental clicks (near a navigation element, close to a scroll path)?

5. If your sample was small (under 200 views): Treat the CTR as directional. Follow up with 5–10 interviews of people who clicked (or didn’t) to understand why. A small-sample fake door plus qualitative interviews is often more useful than a large-sample fake door alone.