Unknown Facts About A/B Testing In SEO

What Is Covered

In my previous article, I have looked at some of the facts that you need to know about web designing. In this article, I want to look at all that you need to know about A/B Testing in SEO. Follow me as we will look at this in this article.

What then is A/B Testing?

A/B testing means comparing two different elements of a website to determine what gets more clicks, conversions, and engagement.

This could include:

oPricing: lower versus higher sales

Offers: a percentage discount versus a set amount off discount (could equate to the same or different values)
Colours: e.g. does a patterned background make users buy more than a grey background?
Placement of CTAs on the page
Timing of animations or pop-up banners: 5 seconds versus 10 seconds

Here we see several examples for where A/B testing can be used:

CTAs: For example, a green CTA box versus a red CTA box, or different offers, such as 50% off versus $50 off for a $100 item.
Headlines: You could write two different headlines for a blog post and test both of them to see which one gets more click-throughs.
Graphic elements: You could test illustrations against stock imagery and see which works best for your audience.
Wording & content: Which works better for your audience, a long landing offer page or a short and concise one?
Type of promotion or offer: For example ‘Buy one get one free’ versus ‘$50 off when you spend over $100’.

Some CMS, like WordPress, have plugins that will do the A/B testing for you or there are tools like Optimizely that you can set up to run an A/B test on your site.

Remember: It’s important to only test one element at a time!

A/B testing can be ineffective if too many tests are running at once: running more than one A/B test on a website doesn’t give us a clear picture of what actually worked.

Be sure to only run one test at a time:

Running more than one A/B test on a website doesn’t give us a clear picture of what actually worked.

A/B testing plugins for WordPress (for WordPress sites only):

NelioABTesting.com: good for testing different elements on a WordPress site
SimplePageTester.com: focused more on testing landing pages

Other tools that can implement A/B testing include:

Optimizely: works across several different types of websites
Five-second test: only offers initial impressions (the first 5 seconds) to gather users’ first impressions
VWMO- test duration calculator: calculates how long to run a test

There are more examples below in the resources. Whatever test that works best for you depends on your CMS, budget, and what you want to test.

A/B testing WordPress plugin examples:

NelioABTesting.com
SimplePageTester.com
https://conversionxl.com/blog/ux-research-and-ab-tests/

More A/B test tools:

https://www.optimizely.com/
https://support.google.com/analytics/answer/1745152?hl=en
https://fivesecondtest.com/
https://wordpress.org/plugins/wordpress-ab-theme-split-tests/
https://vwo.com/ab-split-test-duration/

#1 Pick one variable to test: this could be something like CTA button colour, page layout, campaign offer (e.g. $50 versus 10% off, even though they are the same value).

#2 Choose your goal: what do you want to accomplish with this test? Choosing a specific goal will help you determine whether or not making these changes permanent will make a difference. An example of a goal would be: “Increase conversions 5% in the next 8 weeks.” Make sure it is specific and actionable.

#4 Set up your control and your challenger: only test one thing at a time, and always have a “control,” where the element doesn’t change. This is so you can make sure that a specific element actually makes a difference or not. For instance, if you wanted to test your offer, you would keep your “control” as your existing offer, say, $50. That would be the “A” in the A/B test. Then, you test the “B” part of the test, the challenger, like 10%.

#5 Split your sample groups equally and randomly: you must have the same sample size in order to have a proper test. If one side was bigger than the other, then it may have the advantage.

#5 Determine your sample size: figure out how big your sample sizes are going to be. This is dependent on what you’re testing. For instance, if it was through email, the number of email subscribers helps determine the sample size (e.g. If you have 1000 email subscribers with a 4% click-through rate on average, you can assume 40 people will click through to the landing page you’re testing. This would mean 20 would be shown test A (the control) and 20 would be show test B (the challenge), on average. In most cases, the A/B test software randomizes or alternates it.

For instance, if we tested both the promotional offer and the headline of the landing page at the same time, and at got a higher conversion rate, we don’t know if it was the headline OR the promotional offer that actually caused the higher conversion rate.