A/B testing (also known as split testing or bucket testing) can be an important part of your internet marketing plan. A/B testing involves making a single change to your landing page, newsletter, or blog post. Then, your original AND the altered version are sent out to your clientele. These two versions, A and B, are sent to different lists. If you have a mailing list of 1000 people, 500 will get version A and 500 will get version B. Afterward, you’ll determine which version was closer to meeting your needs.
Check out the Wikipedia definition here.
What Is The Goal?
The goal is up to you. Perhaps you want more traffic to your site. If so, then you’ll want to try things that will increase traffic such as altering the look of the call-to-action prompt in your newsletter. If you want to increase sales, then you might try altering the way the product is displayed or talked about. Maybe you’re just trying to get more people to open your newsletter. In that case, you might try changing your e-mail subject line or the name of the newsletter.
Once you decide what it is you want to achieve, you can better chose what it is you alter and how you deal with it.
What Can I Alter?
The list of things you can alter is huge, ranging from the font you use for your newsletter to the color scheme of your website. Again, what you alter will depend on what your end-goal is. The key is to only make a single change so that you know what variable contributed to your test. Here is a small list of some of the things you could change and what effect they might have:
E-mail Subject Line: Let’s say you the same newsletter to two different groups. You send to 500 people with the subject line “Summer Newsletter.” This is your standard type of subject line, or A. To the next 500 people you send the same newsletter but with the subject line “John Bizowner Makes Summer Better,” which you’ll consider B. The first has an open rate of 18%, while the second has an open rate of 43%. Now you know that, next time, you should send a more engaging subject line to all 1000 of your mailing list.
Landing Page layout: You may want to increase click-through on a particular landing page, such as having more visitors opening your order form or signing up for your newsletter. Making two versions of your landing page will allow you to see what is the most effective layout. Change where you put your offer, such as on the left or right side of the screen. You can add an element, such as including testimonials in one version and not in the second. Does the color of the button make a difference? You can find out with A/B testing.
Forms: If you take the time to create a form, then you obviously want more people filling it out. Alter the layout of your forms and send two versions. That way you can see what color people respond to more, how many questions they are more likely to answer, whether it makes a difference whether it’s anonymous or not, etc.
Web Pages: Honestly, this list is huge. You can test everything from the title-tag of your page to the color of the page itself, from the font size of your text to the inclusion of background music. Anything is fair game!
Start with something that will make a difference in the beginning of your campaign. There’s no point in testing how many people open your page 3 Order Form if you aren’t getting enough people to your site in the first place! Start with the things that will make the biggest impact. Later, you can work your way deeper.
Then, your original AND the altered version are sent out to your clientele. These two versions, A and B, are sent to different lists. If you have a mailing list of 1000 people, 500 will get version A and 500 will get version B. Afterward, you’ll determine which version was closer to meeting your needs.
A/B testing can be a slow process, but you want to make sure you don’t ruin your results by doing too many things at once. If you change the location of a button AND add a new photo, how will you know which one made the difference? And while we’re on the subject of time, here’s an additional tip: Don’t rush your test. Be sure you give it plenty of time. Maybe it’s a week, maybe two, or maybe it takes a month. The length of time will depend on the test you’re running. Take your time!
A Never Ending Loop
Once you have the results of your test, start another one immediately! You want to continually improve your website, and testing is the way to do that.