What is A/B Testing & What Does it Mean for Your Landing Page?

A beginner’s guide to landing page A/B testing and tips on how to test like a champ

As we have discussed in other articles, there are various types of landing pages  and countless techniques for how to create them. But how do you tell what actually works?

Creating the best landing pages isn’t a guessing game like many businesses believe it be. By bringing scientific A/B testing into search engine optimization, we’re able to determine exactly what works for your unique organization and repeat that process to boost maximum ROI. This is where content marketing meets the scientific testing method.

Continue reading to learn about what A/B testing is, what to know before testing, how to run an A/B test yourself and much more.

The ABC’s of A/B Testing

A/B testing is when you experiment with two or more pages at the same time, comparing the data to see which page performs best for your business. Although the process is called A/B or “split” testing, it can be performed on as many pages as you need.

To begin A/B testing, first you must determine the areas of your landing page that you want to test. Then, you will need to create new versions so that you can experiment with these elements. Areas of your landing page that may benefit from A/B testing include:

1. Call to Action (CTA) – How does form length affect conversions? What is the best location for the button?
   
   
2. Video – Does adding a video help or hurt your conversion rate? What if you move it higher up on the page? What if you use autoplay functions?
   
   
3. Headlines – Which page title attracts the most traffic and result in the best conversion rate?
   
   
4. Images – Which pictures are used on the best performing pages? Do certain images negatively impact conversion rate?
   
   
5. Copy – What content scores highest? Does longer or shorter copy perform better? Is a casual or professional tone best?
   
   
6. Layout – Does a single-column or multi-column page result in the most conversions? What if you target the page to mobile users vs. desktop users?

How do you decide the traffic weight during an A/B test?

Before you begin A/B testing, traffic should be assigned to each variant randomly at a predetermined weight. Two suggested weights are 50/50 and 60/40, but the weights could be basically any weight you’d like.

Once you’ve had a visitor land on one of your pages, they should be “cookied” – meaning they’ll land on the same page each time they visit (ensuring the reliability of the test). The number of variants and new ideas on a page will have an effect on the weight you’ll want to assign.

Typically, A/B tests are categorized into two types: multiple variant testing and pre-existing page testing. The type you are testing may help to determine the best traffic weight.

Multiple variants testing

For an A/B testing done with multiple variants, you’ll likely want to give each page an equal weight. This is because you want to be able to pick a champion page based on equal factors to determine the winner. Landing pages created from scratch do not yet have statistically significant data yet, so starting the experiment from equal ground makes the most sense.

Pre-existing page testing

When you’re experimenting with elements of pre-existing pages, you may want to create a new landing page to test these ideas before incorporating them into your control page. That way you don’t apply ideas that don’t end up working. In this scenario, you may want to give the new page a smaller ratio of traffic than the current champion.

Wait! What You Need to Know Before Testing

Now that you know what A/B testing is and what type of test you want to run, you may be eager to get started. However, you should never test your landing pages without having a clear purpose. At the very least, you should have a hypothesis before deciding to start an A/B test.

Before running your A/B tests on actual customers, it’s also a good idea to have some idea as to whether your changes have a chance of working. After all, you don’t want to risk losing customers on an idea that was doomed from the start.

Fortunately, you can get a general consensus of whether your idea could prove successful by collecting preliminary feedback from your site visitors.

Before you know why you are going to run a test, get some actionable insight into what could make your test more effective by:

1. Surveying visitors about their experience on your site. What was done well? What could you do better?
2. Organizing a live chat and speaking to site visitors directly.
3. Installing usability tools that will show when and where people leave your page.
4. Using heatmap software to show what elements are most often clicked on your page.

Do you have a clear purpose for your test? A hypothesis? Is there adequate support for the test? If yes, then we’re ready to move on to one of the most important steps…

How to Test Your Landing Page

Once you’re ready to test, you should remember the following rules to make sure that your experiment is free of errors and as accurate as possible:

1. Ensure that at least 100 different people (unique visitors) visit your test page.
2. Run your test for at least a week.
3. A/B testing requires that you test both pages at the same time so that you can discover trends in daily behavior.
4. In terms of statistics, the significance of the result should be more than 95%—any less could be contributed to chance.

Step 1: Create a landing page (control).

Step 2: Create a page variant.

Step 3: Name the variation page.

TIP: When you’re testing pages against each other, it’s a good idea to label them so you know what’s what. Here are some of the names we recommend:

1. Champion page – This is the winning page at the end of a test.
2. Challenger page - After you have a champion page, you can always run further tests against it. These pages can be called “challenger pages.”
3. Page variant - All pages that you’ve made any changes to are called “variant pages.”

Step 4: Edit the page variant by changing the certain elements you wish to test (CTA, image, copy, headline, etc.)

Step 5: Analyze the A/B test results by several metrics (conversion, traffic, audience, etc.) to determine a winner. 

Ready, Set, Test!

When done right, A/B testing is a fantastic way to measure the success of your landing pages using data-driven results. By incorporating the scientific testing method into your content marketing campaign, you can discover the recipe for landing page optimization and maximize the return on your investment.

Need help with your company’s landing page optimization? SEO Advantage® will help test your website to determine how we can improve your search engine visibility. Our results-driven approach makes us unique and is the secret for getting our clients to rank in the top search engine positions. Contact us to learn how we can assist you.

About SEO Advantage^®

Led by Stone Reuning, SEO Advantage®, Inc. is an online marketing and search engine optimization firm that helps businesses turn their web presence into a powerful revenue generation medium. Our clients enjoy dominance on Google, Yahoo and Bing through a suite of unique pay-for-performance search engine optimization and online marketing services implemented by a multidisciplinary team of SEO engineers, copywriters and web designers. You'll find us referenced in books such as "Writing Web-Based Advertising Copy to Get the Sale" and the BusinessWeek bestseller "The New Rules of Marketing & PR," as well as the popular ebook "The Small Business Blogging Blueprint." Find out more today by calling us at 1-800-366-1639 or email us here.