Experiment Examples To Get You Started Testing Your Idea

Robbert van Geldrop
August 11, 2016

One of the most asked questions I come across when coaching and advising startup founders are about examples. I decided to create a collection of the best example per category and give you some inspiration for your next lean startup experiment.

To get the theory out of the way of this post, I’d like to refer you to Tristan
Kromer’s blog post about how to structure your experiments


Uber founder Travis allegedly cold-called ten black taxi driver and pitched them about the service. 3 out of 10 wanted to set a meeting with him after this call, which is exceptionally high for a cold-call conversion rate.

Kien.io — Dutch startup from Delft interviewed customers on how they experienced their existing wireless sound systems. One might say that this is not the preferred way of testing in Lean Startup since it doesn’t relate to finding problems. Kein.io showed that using their research they understood better the minor annoyance people had with wireless sound systems. The learnings allowed them to improve the design of their product and market it effectively.

Eternal Sun is a Dutch start-up that developed Solar Cells. To test their products, they developed in-house testing equipment. One day a potential customer arrived and spotted the testing equipment. Apparently there was a common problem. The customer bought one of the testing setups on the spot.

All these startups learned through real and in-person customer interactions.
Remember to get out of the building actively, you’ll learn things that you
didn’t have on your radar before.


An interesting example of an experiment comes from Facebook. Back in the days
Mark Zuckerberg started a project called Facemash. He learned that people engage more with people they already know. Opposed to engage with random strangers. The insight is something he would only learn from actual behavior by people joining Facemash. After it’s first inception ‘The Facebook’ was launched, and the rest is history.

Experiments come in many forms and shapes, we’ve categorized a few as an inspiration for your startup.


Buffer is a well-known example. Development of the product too longer then expected. Whilte questioning if people would actually want their product they performed a ‘smoke test’. On a simple landingpage Buffer pitched their product and measure interest.

Tons of startups have done this before. It’s the textbook method of applying Lean Startup, yet it also gets done badly on occasions. It’s not always the best way for learning, but for SaaS startups, this can be a good way to start.


Groupon didn’t start out as a daily-deal or couponing platform. At first they were called ‘The Point’, a group action site for social causes. With their business model not catching on, they tried something new. On their blog, they published an article which offered a group discount to the pizza joint around the corner. The deal was such a success that they pivoted to the model of discounts and daily deals. Read the full story here.


Dropbox still holds the lead here by creating a video which demonstrated the product well before it was finished to capture sign-ups. Note that the riskiest assumption isn’t always related to a pain point of a customer. In existing markets with existing solutions, you might have to assume that other elements are at risk, such as price points or customer channels. Dropbox never tested (or had to test) the problem they were fixing. They only tested a better solution, being automatic syncing between devices.

There’s an excellent talk online from Drew Houston about the experiment:


Please also watch this to learn a valuable lesson about engines of growth as Dropbox pivoted in the way they were acquiring customers.


AirBnB started out as a Concierge MVP. There was a design conference coming to town, and they decided to open up their loft as cheap accommodation for conference attendees who had lucked out on the hotels nearby. They took pictures of their apartment, put it up on a simple website, and soon they had 3 paying guests for the duration of the conference, a woman from Boston, a father from Utah, and another man originally from India.

Image Source: TechCrunch.

More info here.

Airbnb still had to figure out how to scale and did so by discovering that the listings would be booked much faster in case the photographs were of high

An excellent Dutch example is 3D Hubs. This startup initially served their customers by going around in Amsterdam and finding owners of 3D printers willing to take payed printing jobs. Read the full story here in the AMA of one of the founders at Reddit.


This is a similar way of working as the more known ‘concierge MVP’ with the difference that it happens ‘behind the curtains’. Customers are under the impression that they’re using a finished product, while behind the scenes it’s run manually.

Aardvark is the textbook example of a Wizard of Oz. They proposed an algorithm which would find answers to your questions through social media, but in reality it was their team doing all that work manually. They were acquired by Google. There’s a video about it:


An experiment where you use a competitve product to test your assumptions with. The most famous example is Zappos. Founder Tony Hsieh simply bought the shoes he offered online in local retailers. Full story

One of BackupAgent’s Dutch competitors unintentionally performed an Imposter Judo experiment as they initially were a reseller of another backup software vendor. As such they were able to validate the demand for the service as well as the shortcomings the existing software had in order to facilitate those customers well.