What Is the Mechanical Turk Pretotype?
The Mechanical Turk pretotype presents what looks like a fully automated product, but a hidden human secretly performs the core function behind the scenes. You test whether people actually want the outcome — and how they react to the experience — before building any of the real technology.
Quick Answer: A Mechanical Turk pretotype fakes automation with hidden human labor. The user sees a slick "it just works" product; a person behind the curtain does the work a machine would eventually do. It validates demand and experience fast, cheaply, and without writing the hard code first.
The name comes from an 18th-century "automaton" — a chess-playing machine that beat opponents across Europe until it was revealed to hide a human chess master inside. The pretotype borrows the trick on purpose: look automated, be human. It is one of the core techniques Alberto Savoia describes in The Right It, and it answers a single, decisive question before you invest in engineering — do people actually want this?
How the Mechanical Turk Pretotype Works: A Human Behind a Fake Machine
The mechanics are simple: build a convincing front end, wire the back end to a person instead of software, and let real users interact as if the automation were real. The customer never needs to know a human is doing the work — that secrecy keeps their experience honest.
The front end must feel automated. A form, an app screen, an email address, or a chat window is usually enough. Users submit a request expecting a machine to handle it, and the interface should give no hint that anything else is happening.
A human quietly does the core work. Behind that interface, you or a teammate manually produces the output — writing the summary, matching the request, generating the recommendation — that a finished product would eventually automate. This is the same operating principle as a Wizard of Oz MVP: the user believes they are talking to software, but a person is pulling the levers.
Speed and quality still have to look real. Because you are testing the real experience, the human needs to deliver on a believable timeline — if the "instant" result takes three days, you are measuring patience, not demand. This technique sits within the broader family of pretotyping methods that fake the product to test the market before you build it.
A Concrete Mechanical Turk Pretotype Example
The clearest examples pair an ordinary-looking interface with a human doing the labor invisibly. Imagine a founder testing a service that promises AI-written meeting summaries delivered minutes after any call.
The pretotype: users forward a recording to an email address and receive a polished summary back. There is no AI. The founder listens to each recording and writes the summary by hand, then sends it from a branded template so it feels machine-generated.
What that setup reveals: whether people actually forward recordings, how many return a second time, and which parts of the summary they quote or complain about. Every one of those signals is real demand data — collected before a single line of transcription code exists. If almost nobody forwards a recording, the founder has saved months building an engine no one wanted.
The illustration above is a teaching scenario, not a case study, but it shows the technique's shape: the outcome is genuine, the automation is faked.
Mechanical Turk vs Wizard of Oz: Nearly the Same Idea
For practical purposes the Mechanical Turk and the Wizard of Oz MVP are the same maneuver — a human faking automation — and the terms are often used interchangeably. The table below draws the fine distinction people sometimes make, so you can decide whether it matters to you.
| Aspect | Mechanical Turk pretotype | Wizard of Oz MVP |
|---|---|---|
| Core idea | Human secretly does the automated work | Human secretly does the automated work |
| Emphasis in usage | Often the whole core function is human-powered | Often a specific feature is faked inside a broader product |
| Origin of the name | 18th-century chess "automaton" hiding an operator | The wizard hiding behind the curtain in Oz |
| Source | Savoia's pretotyping techniques in The Right It | Widely used across lean and UX practice |
| What it validates | Demand and experience, not feasibility | Demand and experience, not feasibility |
Takeaway: Do not spend energy policing the boundary between these two labels — they describe one behavior. What matters is that a human is standing in for the machine so you can measure whether anyone wants the result before you build it.
When to Fake the Tech Before You Build It
Reach for a Mechanical Turk pretotype when the hard, expensive part of your idea is the automation itself and you are not yet sure the outcome is wanted. Faking the technology lets you test desirability without paying the engineering bill.
Use it when the technology is the risk. If building the real thing means months of machine-learning work, integrations, or infrastructure, faking it first is the responsible move — there is no sense automating a job nobody will pay to have done.
Know exactly what it does — and does not — prove. A Mechanical Turk pretotype validates that people want the outcome and will engage with the experience. It does not prove the technology is feasible, and it does not scale — one human cannot serve thousands of users. Because a person is doing the work, the experience may also run slower than a finished product, so weigh that when reading results. Treat it as a demand test, then graduate to real automation once demand is confirmed. It slots naturally into a wider startup idea validation sequence, where cheap experiments come before expensive builds — and once you have your answer, stop, because running a manual operation indefinitely is a business decision, not a test.
Key Takeaways
- A Mechanical Turk pretotype fakes automation with hidden human labor. The user sees an automated product; a person behind the scenes performs the core function, so you test demand before building the technology.
- The name comes from a fake 18th-century chess automaton that concealed a human operator — the same trick, used deliberately to look automated while being human-powered.
- It is essentially the same idea as a Wizard of Oz MVP. The labels are often interchangeable; both put a human where the machine will eventually go.
- It validates desirability, not feasibility. You learn whether people want the outcome and engage with the experience — not whether the automation can be built.
- It does not scale, and that is fine. One human cannot serve everyone; the pretotype's job is to confirm demand, then hand off to real engineering.
- Use it when the technology is the expensive risk. Faking the hard automation first means you never build a machine for a job nobody wanted done.
- It comes from Savoia's The Right It. The technique is one of several pretotyping tools for collecting your own demand data cheaply, before a full build.
Frequently Asked Questions
What Does "Mechanical Turk" Mean in Pretotyping?
It refers to faking an automated product by hiding a human who performs the work behind the interface. The name comes from an 18th-century chess-playing "automaton" that secretly held a human operator. In pretotyping, you use the same illusion on purpose so users experience a seemingly automated product while a person quietly does the work.
Is the Mechanical Turk Pretotype the Same as a Wizard of Oz MVP?
For practical purposes, yes. Both hide a human who performs the function users think software is handling, and the terms are frequently used interchangeably. Some people reserve "Mechanical Turk" for a fully human-powered core and "Wizard of Oz" for a faked feature inside a larger product, but the underlying maneuver is identical.
What Does a Mechanical Turk Pretotype Validate?
It validates desirability — whether people actually want the outcome and will engage with the experience — collected as real demand data before you build any technology. It deliberately does not test feasibility, because a human is doing what a machine eventually would. Use it to confirm demand, then move to real automation.
Does a Mechanical Turk Pretotype Scale?
No, and it is not meant to. A single human standing in for software cannot serve large numbers of users, and the experience may run slower than a finished product. The pretotype exists only to answer whether the outcome is wanted; once demand is confirmed, you replace the hidden human with the real automation.