Also called: Manual-first, Mechanical Turk
See also: Concierge
Strength of evidence
Relevant metrics: Acquisition, Activation, Cycle time, Customer feedback, Cost
Validates: Feasibility, Viability, Desirability
How: Use your own hands, an intern, or online crowdsourcing services to fake automation of tasks that are now too costly to build. To keep the facade of no human involvement, consider constructing your experiment so that real-time response is not needed in order to deliver on the value proposition you're testing.
Why: Humans can be cheaper than automation. Even if this takes longer for the customer to receive an answer, you will avoid wasting precious time building features the customer does not want.
This card is part of the Validation Patterns printed card deck
A collection of 60 product experiments that will validate your idea in a matter of days, not months. They are regularly used by product builders at companies like Google, Facebook, Dropbox, and Amazon.Get your deck!
Before the experiment
The first thing to do when planning any kind of test or experiment, is to figure out what you want to test. To make critical assumptions explicit, fill out an experiment sheet as you prepare your test. We created a sample sheet for you to get started. Download the Experiment Sheet.
Creating an illusion to prope demand
Using the Wizard of Oz experiment, you are creating an illusion of an actual intended product, where the unsuspecting user will never know that you are in fact handling all suposedly automated tasks, manually. From the point of view of the customer, your Wizard of Oz prototype will do the job in a way that end-users are not aware that humans are working behind the scenes.
Why would you do this?
First, a Wizard of Oz prototype provides a unique opportunity to verify the demand of your product that, so far, only exists in your mind. So why spend time and money building an elaborate system to handle and automate your customers’ requests, when you can just test whether anybody are interested at all in your future product, by hand-holding the entire process yourself.
Secondly, conducting all tasks manually will provide unique insights into what it takes to deliver customer value as well as how users react when you finally deliver value to them.
A flexible solution to adjust as you learn
Wizard of Oz testing is a fast and effective way to test your hypothesis as to whether your proposed solution will in fact create value for your potential customers. At the same time, performing all tasks manually comes with the added benefit of being easily adjustable. Being able to quickly modify your hand-held product experience lets you test a larger number of hypotheses, quickly – finding the most effective solution, fast.
After the experiment
To make sure you move forward, it is a good idea to systematically record your the insights you learned and what actions or decisions follow. We created a sample Learning Sheet, that will help you capture insights in the process of turning your product ideas successful. Download the Learning Sheet.
The tools below will help you with the Wizard of Oz play.
Zappos founder, Nick Swinmurn, decided to test the assumption that people would be willing to buy shoes online without trying them on. Instead of building an inventory and then trying to sell it, Nick took a different approach. He went to local stores, photographed shoes, and then advertised them online. If a pair of shoes were sold, Nick would go back to the store, buy the shoe, and send it to the customer.
From the customers perspective, it would look like Zappos had a full inventory of shoes, while in fact, all orders were processed in this unscalable way manually. Nick could use this experiment to fine tune marketing, wording, packeting, and product categories in terms of what constituted better product/market fit. Once momentum was gained, Zappos could push play and invest in expensive inventory and handling as the certainty of success was maximized.
Source: The Wizard of Oz MVP
The Q&A service, Aadvark, routes questions to expert users (via instant messaging). In the early days, the Aadvark staff would manually post the questions to whomever was online to see if that particular user would respond, and then manually post the answer back to the asker. There was no automation or algorithm. The feature was later built to automate this functionality. From the outside, this MVP looked like a fully functional system, even though most tasks were manually executed behind the scenes by a human.