SendGrid Email Marketing Service Tutorial: How to A/B Split Test Jillian Wohlfarth August 26, 2013 Best Practices, Product, Technical // SUMMARIES ?> We’re happy to continue our VidGrid tutorial series in support of the release of our Marketing Email Service. This series will walk you through each step of creating your email marketing campaign to make sure that you have the best experience possible. In this video Kayla provides simple instructions for how to A/B test your marketing email. This feature allows you to create multiple versions of your marketing email, that based on recipient activity, automatically determines your winning variation and then sends it to your remaining recipients. Video Transcription Have you wondered how to create the best newsletter that will be the most well received by your recipients? A/B split testing allows you to do just that. With A/B testing, you are able to create multiple variations of your newsletter and send out those tests to a sample population of your recipient list. Once those test variations have gone out, you’re able to choose a winning variation and that variation will be sent to the remaining recipients on your list. How to Send an A/B Test With SendGrid Let’s take a look and see how to schedule an A/B test to send. From your newsletter dashboard, you will click to set up a newsletter just like normal. On the first set up page, you must first give your newsletter a title. To the right is where you can choose to split your test. Right here where the option says “Split test my newsletter.” You’ll want to select that option. Once you select that option, you’ll see that there’s a scrollbar that allows you to define the size of your test population. By default, our system sets the test population to 30% of your recipient list. That means that 30% of your recipient list will receive that test and the remaining 70% of the recipient list will not receive the newsletter until the winner is defined. You are able to change that in 5% increments. We’re going to use a test size of 35% for this example. Onto Step 2, we will select our recipient list. You can either create a new list or select from existing lists that you already have. I’m going to select my list and move on to Step 3. Design Different Variations of Your Newsletter Step 3 is where you actually design different variations of your newsletter. Once my template has been selected, I’m able to modify the content of the template. On the right here you can see that 2 variations are shown by default. Since this is an A/B split test, there will be at least 2 variations. You are able to create up to 10 variations. If you decide that you’d like to add an additional variation, you simple click the “add a variation” button. Once you click the “add a variation” button, you are given 2 options. You can either use a duplicate of the current newsletter that you’ve created or it will provide you with an empty variation. Now we are able to edit the variations, so let’s start with Variation 1. Oftentimes, having a different subject can help distinguish between your variations. I’m just going to add a basic picture to our example here and move on to Variation 2 and give this variation a subject as well. In this example, I’m using the same content for each of my newsletter variations and only distinguishing different subject lines to determine which one will be most well received. Now onto the third Variation. Schedule to Send Your Variations Once you have edited your variations and you have the content that you like, you’re able to schedule to send the variations. That is Step 4. As you can see from the top of this page, our test is being sent to 35% of our recipient list. Or 6 recipients. That means that 2 recipients are going to receive each variation. You can scroll down to see what each variation looks like. There is a drop down option here that will allow you to switch between the variations. Once you have reviewed the content of your variations, you are able to scroll down to the bottom of the page to schedule your send. You can schedule the send for a later time or you can send the test immediately. I’m going to choose to send the test now. Once I choose to send the test, there’s an option below on how to pick the winner. You’re able to manually choose the winner in which case you would have to come back to the newsletter app to select the winner yourself and it would go out, or you can have the winner be chosen on unique opens or unique clicks. I’m going to choose unique opens to be my winning determinate. Then you’re able to specify how long you’d like between your test and your actual send to go out. I’m going to choose 24 hours. That means that my test will go out now and after 24 hours the winner will be determined and the remaining recipients will receive that winning variation. Now I’m going to finish and send my test variations. If you decide to manually choose your winner or you’d like to send to your winner before your specified time period, you’re able to do so. If you go to the statistics for your split tested newsletter, you are able to review those statistics and pick a winner. You can see at the bottom of the statistics page, it gives you the option of the different variations you’ve created and you have the option to send now. And that is how you create and send your winning variations! To read all of our VidGrid tutorial series so far, click here or you can access the whole library here.