MVP #4: Fake door test

Fake door MVP (also called “Painted door test”) allows you to test market demand for a product or service that doesn’t yet exist but you are considering developing it. It invites people to try or buy the product with a clear call-to-action button to do so, but this button only takes them to another page that reveals the truth – the product does not yet exist. You are measuring the interest by observing the percentage of people who clicked the CTA button.

You can also prod users into leaving you their email addresses to get notified once the product is launched or attend beta testing. This is a much stronger indicator of interest than CTA button clicks, and you’ll also get a pool of potential new users who you might get feedback from during the product development.

What to test: Product or service ideas

Cost: Medium, you need to build a landing page and run ads to get relevant traffic. Alternatively, you might leverage existing audiences using Email campaigns.

Evidence strength: Strong, you’re getting data on what users actually want to do, not just what they say they would do.

Metrics: Clicks on call-to-action buttons, email subscriptions

One of the best examples of a Fake door test is the story of how Buffer was founded:

How Buffer emerged from a fake landing page

Testing interest vs. willingness to pay

In the Buffer example, founder Joel Gascoigne first tested the interest in his product idea by creating a fake landing page describing its key benefits. He then run online ads to drive traffic to the landing page and observe what percentage of this traffic clicks the CTA button.

Once he validated that there were enough people interested in the described value proposition, it was time to test whether they were also willing to pay for it.

Joel created a pricing page where he introduced 3 plans with different features and limitations – free, $5, and $20 in monthly subscription costs, and observed which plan would be clicked on most. Most of the clicks went to the free plan, but around 5% chose the $5 plan.

Although these numbers might not seem like out of this world, it was enough of an indicator there is an audience willing to pay. Also, this performance could be increased by learning who these people are and targeting them better in future campaigns.

Honest or not?

You have two options – first, be honest and admit that the product is still in development, asking visitors to pre-register for it.

The second option is to create an illusion that the product already exists and only tell the truth once people click the CTA button. You’ll get more data this way because clicking a button is much less of a commitment than pre-registering for something, but some visitors might feel intimidated. To avoid that, you should choose your words carefully when revealing that the product is not yet ready.

The CTA button might be something like:

• Get started
• Sign up
• Register
• Pre-register
• Contact sales
• See plans

Testing multiple value propositions

If you’re unsure what your potential new product should look like, you can Split test multiple variations of your landing page to test:

• key benefits
• messaging
• positioning
• visuals
• price

If you’re using online ads to drive traffic, all variations should be accompanied by appropriate ad messaging.

In Kontentino, we created a fake door test to determine whether our new planned product for social media ads management should include only Facebook and Instagram platforms, or the market requires more (e.g. Twitter, Linkedin or TikTok).

We created two fake landing pages with different messaging accompanied by appropriate Facebook ads and AB tested the results to determine which version would yield more email addresses.

Testing marketing performance

Another way to leverage Fake door tests, especially if you’re planning on using online ads to drive growth, is to estimate your ROAS (Return On Advertising Spend), calculated by dividing the LTV (Life-Time Value) by CAC (Customer Acquisition Cost). 

First, let’s calculate your CAC. Let’s say it costs $100 to get 1000 visitors to your page, 0.5% will choose one of the paid plans. This means that for $100, we can get 5 customers, and therefore a single customer is expected to cost us $20.

Now the LTV. This is your sales price. If you run a subscription model, you need to estimate retention first. Let’s say your primary plan will cost $5, and you expect an average of 12 months of retention. $5 times 12 will give you an LTV of $60.

$60 divided by $20 will get you a ROAS of 3, which means that for each dollar spent on advertising, we’ll get 3 back. That is a pretty good ratio, and now we know that the product can be effectively grown using online ads.