introduction to a/b split testing

An Introduction to A/B Split Testing

affiliates Advice, Featured Leave a Comment

Unfortunately, increasing your revenue as an online publisher is never just a question of increasing the number of advertisements or expanding your blog’s advertising space. Instead, it’s about trying multiple blog configurations in order to find the optimal placements for those advertisements. However, how does a blogger go about finding that all-important layout where their click-through rate (CTR) is optimized? The answer comes from using an A/B split test. So, what is an A/B split test and why should all bloggers and affiliate marketers use this blog layout optimization tool?

An A/B split test simply involves comparing one controlled blog or website layout versus another modified layout where one single change has been made. In fact, it’s essential that only one change is made at a time. If you make too many changes to your blog’s layout at one time, then you are never able to define which specific change lead to a rise or decline in CTR. Instead of doing a methodical test, you’ll be left making random guesses as two which change increased your blog’s CTR and revenue.

The proper approach is to make a single change and then track its impact on CTR. Most bloggers focus on measuring the same amount of traffic from week-to-week or month-to-month. For instance, if your blog has 25,000 page views a month, then it’s best to perform one single change at a time and then leave it as is for the entire month before making an additional change. You track each of your changes and then measure the impact on CTR at the end of each month. A good strategy is to use an excel spreadsheet where you can keep notes and track your monthly CTR percentage.

An Example of an A/B Split Test

Let’s assume your original layout has a link advertisement just above your blog’s customized banner. Your current CTR is 1.20 percent, but you think you can improve on this. Therefore, you decide to position that link advertisement just below your navigation bar. After making this single change, you see your next month’s CTR increase to 1.30 percent.

The month after that you change the font on that link advertisement to a more appealing color. In doing this single change you see your CTR increase to 1.45 percent. The next month, you move another advertisement up closer to the top of your blog near your “follow me” section. That change leads to yet another increase in CTR as you achieve 1.55 percent. However, the next month you decide to move that same advertisement to the left side of your blog and your CTR declines to 1.30 percent. In this case, this is not a change worth pursuing.

An A/B split test could easily be referred to as an A/B/C split test. This is because a number of blogs split their traffic equally among three layouts in order to find that optimized configuration. However, this tends to be done by blogs and websites with extremely high traffic volume of 1 million monthly visitors or more. As such, most bloggers do the A/B split test week-by-week or month-by-month. Again, do one change at a time, measure the change in CTR and always make sure you’re comparing the same traffic volume on a monthly basis.

affiliatesAn Introduction to A/B Split Testing