Produced by Jason Zagami


Are you frustrated with low conversion rates on your website, emails or digital ads? Everyone’s been there. There’s an easy fix I’m going to show you in less than 5 minutes.  

Click to Learn How to Run Your First A/B Test

Today you’re going to learn how to Optimize Conversions like Netflix without Hiring a Single Analyst. 

I’m Jay Zagami and this is the CPM Podcast, brought to you by Seedling. 

Netflix constantly tests and optimizes everything, including TV & Movie artwork, that’s the card you see with an image and title. Through testing, they were able to significantly increase the number of plays by A/B testing artwork card segmentation. In their case study, they show how it works using the movie Good Will Hunting as an example.  

People who watch more comedies are more likely to watch Good Will Hunting when the artwork shows a scene with Matt Damon and Robin Williams, a comedian.  

People who watch more Romantic films are more likely to watch Good Will Hunting when presenting with artwork card showing a romantic scene between Minnie Driver and Matt Damon.  

They tested this against  

  1. Pre-chosen artwork (control) 
  2. Randomly chosen artwork   
  3. Contextual Chosen Artwork 


Image Credit

The contextual artwork showed a significant uptick in views - their chosen conversion.  

This is a more complicated example that what you need, but everyone knows Netflix so It’s relatable.  

If you’re not constantly testing and measuring conversions, at a minimum, you're missing out on a huge opportunity to generate more revenue. At worst, you’re letting your competitors run away with your customers at a speed you most likely won’t catch up to. 

So here’s how to run powerful A/B tests in 3 easy steps.  

You can do this on an email, a web page or digital ads. Most email service providers natively include the ability to create variants for A/B testing. Facebook and Google allow you to run variants of ads for A/B testing, and there are easy ways to run tests in WordPress, Squarespace, and HubSpot. Each of these services includes some form of data reporting so you can see what happened.  

Here it is: 

Step 1. - Choose an element, an image, a headline, a conversion offer. Doesn’t matter what it is, but only choose 1 thing. Come up with a hypothesis around how you could improve that element. For example, your form button may say “Submit” what if you tested one that said “Send” or “Sign Up Now.  

Step 2. - Create the variant and run it to statistical significance. This means that there is enough data to be sure your test is accurate.  

Step 3. - Look at your data and pick the winner. Make sure 100% of your traffic goes to the winning option.  

Once you have a winner, start planning your next A/B test.  

Click to Learn How to Run Your First A/B Test

If there’s a topic you want me to cover in future episodes let me know on twitter @jasonzagami. I’ll see you next time.