Get Mentorship on Experiments from Adam Berk

Adam is an adept expert of LEAN startups and running experiments. To him, the only thing that matters is traction. Finding the product market fit is the goal and to get there you have to test, test and test.

Make one person happy and understand why it’s creating value for this individual. You don’t know how you did it until you do it again. The second you do that everything changes.

Revenue, retention and referral are the proof that what you are doing is working and appealing to the customers. There is no better evidence than people coming back and referral is social proof, it’s currency.

Start small, prove it once and then grow. The way to do this, I think you guessed, is to run experiments. Ok so what are the components of a good experiment? It has to be based on a quantifiable, timed and testable hypothesis. Speed is important in the process, you have to go fast and test it quickly. There is always a way to test a feature even if it’s not scalable and if it’s critical to your business don’t wait. Start with the very risky things, it’s better to do it now than in 6 months when you’ll have spent a lot of money on your project. There are a few cool hacks to isolate variables and bend time. It’s not sustainable (programs do things that don’t scale) but it’s a good way to isolate components and test them independently. Make sure you hold some variables constant while you test the riskiest ones. You have to isolate them for the test to be efficient. Test everything you don’t agree with with your team. Get real numbers back and build on that. When you have science on your side the backlash is always minimized.

“You want to test things that are the most unknown and most crucial to your business model.”

“You have to be willing to do certain things for the sake of learning” You are the one to make your ethic calls and see what is acceptable. Don’t run test to annoy people but if a few get annoyed it’s ok, it’s how you learn and improve and sometimes it’s necessary to have some negative feedback. If you’re not too sure you can keep it contained by testing it on a subsample until you have a better solution.

The examples of payment and pricing

To test payment put a paypal button (or even put a button “click here to pay”) and observe your users’ reactions. Understand why they don’t want to pay by asking them directly.

According to Adam, pricing is the most fun to test. You can test price all the time and sometimes you’ll see that more people pay when the price is higher. It’s not because you set a price that it never can change again.

“Perfection is achieved not when there is nothing left to add but when there is nothing left to take away”

Let us know what you tested in your business and how in the comment section.


─ March 31, 2016

Signup to our Newsletter!

Subscribe to stay connected with our latest news