Email Marketing: How to run A/B Tests within Klaviyo (Step-by-Step Email Marketing Tutorial)

Hey guys, so for this video, I'm going to dive into how to do an AP test within Klaviyo. I've been training kind of my brother, some other people on the team and everyone's been asking about how do I run an AP test. So depending on where you're at in your journey with email, I figured this could be either really useful all the way to somewhat useful. So let's go through it. So I have a campaign ready. I'm going to send this campaign to a 30-day engaged. There's about 8,000 people. So I already have the content ready and whatnot. I'm going to go over here to save and continue. And when you're here, right, you've got the from name, from email, the subject, the preview text, and then all the content, right? Once this is all done, we're basically going to go over here to click on where it says add variation. So now you can see that we have version A, version B. So in the case that we want to do a subject line test, all we have to do is go over here to edit variation and then change the subject line. Right now it's time to stop sleeping bacteria. You know, again, this isn't super creative. We'll want to change this. I'm going to be different, better. Let's just go with this for right now. We'll press save. And then we'll just go back here to message content. So we'll see now that version A has this subject line. Version B has this subject line. They're both weighted 50-50. So now here's how we do the test settings. So because the list was 8,000 people in total, we typically want about 1,000 people or more to receive each variation. So we'll make sure that, you know, 25% of 8,000, right, should equal each one being right at or above that number. So 8,000 times 25. That's almost perfectly, yeah, that's perfectly meaning that each person is going to get 1,000, or each variation is going to be 1,000 recipients. Again, the more people that we can do, the better. So we'll leave this at 25%. You can easily change this to 35, you know, any other number, right? So we'll leave this at 25%. How long should we wait? This is also a question that you get. You want to wait long enough that you get enough data without waiting too long if the email is time sensitive. So typically, I'll realistically test anywhere between, you know, maybe two to four hours, with about three hours being the most common and popular. But again, like, I'm also optimizing for some time and other things to keep in mind. But say you want to just do it for three hours, right? So, you know, 25% of the people are going to get the A-B test. So 12.5% will receive version A, 12.5% will receive version B. And then in this case, because we're testing the subject line, we want this to be chosen based off the open rate. However, if we are testing, you know, an offer within an email or plain text versus HTML or something else, then we'll want the subject lines to be the same. And we'll want something within the content to be a little bit different. In that case, we would just test off of unique clicks. So we'll just go here, save and schedule campaign. And then you'll see, again, like version A here, version B there. All you have to do is go to schedule and send. And that's pretty simply how you do, you know, an A-B test on campaigns. I'm going to do, you know, we can do actually one really quickly for flows while we're here. Let's see. All right. So let's just say in the browse abandonment email that we want to do an A-B test. In this case, there's already one that's running live. But I'll go ahead and just create an additional one. So we'll go over here. And basically what we'll do is we'll, you know, add a variation. Say we want the content to be the exact same, which we would. Here, we'll basically just go to duplicate variation. And then, you know, pretend that we only have two. But if we had these three, we wanted to run three tests. All we have to do is click equalize weights and we'll go to one third piece. But for here, again, if we just want to edit this, and this is just going to be a subject line test, you do the exact same thing, right? New subject line here. Changes. Go to done. And now we have three different variations running. So, yeah, just pretty simple and easy to do. Before I forget, I'm just going to delete this out. Delete variation. And that's exactly how you can do a subject line test. If this happens, again, all you have to do over here if you want it to be 50-50 is equalize the weights, but also just type it in directly there and hit done. So, that is the lesson on A-B testing. As always, if you have any questions, feel free to reach out. Thanks, guys.