Pretotyping is a set of techniques for quickly and cheaply testing the viability of an idea before building a full prototype. The term was coined by Alberto Savoia, a former Google engineer, combining "pretend" and "prototype".
Pretotype ≠ Prototype. A prototype answers the question "Can we build this?". A pretotype answers the question "Should we build this?"
Why Pretotyping Matters
Most new products fail not because they are poorly built, but because nobody wants them. Pretotyping lets you test demand in hours or days instead of months of development.
- Save time and money on products nobody needs
- Get real data instead of opinions
- Iterate on ideas quickly
- Reduce the risk of failure
8 Pretotyping Techniques
1. Mechanical Turk
The user thinks they are interacting with an automated system or technology, but behind the scenes everything is done by a human. This lets you test demand before building complex technology.
Nick Swinmurn created a website with photos of shoes from local stores. When someone ordered, he personally went to the store, bought the shoes, and shipped them. This way he validated whether people would buy shoes online - without building any logistics infrastructure.
When to use: When you need to test demand for a service you plan to automate. It also lets you deeply understand customer needs.
2. Fake Door
Create a landing page or button for a feature that doesn't exist yet, and measure user interest by the number of clicks.
Drew Houston created a video showing how Dropbox would work. Overnight, the waitlist grew from 5,000 to 75,000 people - before a single line of real product code was written.
How to apply:
- Create a landing page describing the product
- Add a signup form or "Buy" button
- Run ads targeting your audience
- Measure conversion (how many people clicked/signed up)
3. Facade
Unlike Fake Door, here you actually deliver value to the customer, but you do it manually without automation. The customer gets what they paid for, but you haven't invested in development yet.
A major airline ticket seller wanted to test whether people would buy Aeroexpress train tickets alongside flights. Instead of building an API integration (contracts, security audits, months of work), they simply pre-purchased 20-30 tickets (~$70). When a customer clicked "Buy", they were shown one of the pre-purchased QR codes. In a single day they measured conversion and calculated the ROI of the full integration.
4. Pinocchio
Create a non-functional physical version of the product and place it in real-world usage conditions. This helps uncover problems you never even thought about.
When designing the PalmPilot, the developer simply printed a mock interface on a piece of cardboard and tried to use it as a real device throughout the day. This helped understand the ergonomics before writing a single line of code.
When we showed a sleek box for the overhead bin to Boeing executives, they said: "It will be stolen on the first flight!" We hadn't even considered that. It turned out the box needed to look "ugly" and official so passengers wouldn't take it.
5. Video
Instead of building the product, create a video demo and measure audience interest. With the rise of AI tools (Veo, Sora), this technique is becoming more accessible than ever.
When Uber was planning an air taxi service, they created a fully rendered video (not real!) and distributed it on YouTube. This allowed them to gauge interest without building an actual flying prototype.
Distribution channels:
- Kickstarter/crowdfunding - people vote with their wallets
- YouTube/TikTok - measure views and engagement
- Newsletter - build a list of interested users
6. Infiltrator
Place your product in an existing marketplace where your target audience already is. Use someone else's traffic to test demand.
Airbnb's first listings were automatically cross-posted to Craigslist, where there was already an audience looking for housing. This brought in the first users without any marketing spend. Many successful services "emerged" from Craigslist, Avito, and OLX.
A team created a prototype key holder and placed it on a shelf in IKEA as if it were a real product. When people tried to buy it at checkout, they realized it was not an IKEA item. This measured demand in real-world conditions.
7. Small Test Market
Test the product with a limited audience or geography before a full-scale launch. This lets you optimize the product in "safe" conditions.
Facebook started only at Harvard, then expanded to other colleges, and only later opened to everyone. This allowed them to refine the product in a small community before scaling.
During the mobile app boom, all games were first tested in Canada or Australia before launching in the US market. Similar audience, but lower risk.
8. Imposter
Implement only the core technology and place it inside the shell of an existing product to demonstrate the viability of the idea.
When Elon Musk presented the Tesla concept, he took a Lotus Elise body and simply replaced the engine with an electric one. This allowed him to demonstrate the concept and collect $5,000 deposits without designing a car from scratch.
How to Choose a Technique
The choice of technique depends on what exactly you want to validate:
- Fake Door - is there demand for the idea at all
- Mechanical Turk - are customers willing to pay and what service do they need
- Facade - does the business model work and what is the ROI of automation
- Pinocchio - how the product will work in real conditions
- Video - can you attract attention to the idea
- Infiltrator - where to find the first users
- Small Test Market - how to optimize before scaling
- Imposter - does the core technology work
It is often more effective to combine techniques. For example: Fake Door to test demand + Video to demonstrate the product + Infiltrator to find first users.
On Ethics
Pretotyping techniques raise ethical questions. This is a personal choice for each founder. A few recommendations:
- You can honestly tell users the feature is not yet available, but you are gauging demand
- Offer to leave an email for a launch notification
- Facade is more ethical than Fake Door because the customer receives actual value
- Each technique can be adapted to your comfort level
Success Metrics
Pretotyping without clear metrics is pointless. Define in advance:
- XYZ hypothesis: "At least X% of audience Y will take action Z"
- Success criteria: what result counts as validation of the idea?
- Failure criteria: at what result do you abandon the idea?
This article is a brief overview of one of the topics from the "AI Founder" course. In the course, you will not only study these techniques in depth but also apply them to your own idea with mentor guidance.