How To Become an A/B Testing Badass

Do you ever wonder if your website’s landing page is as effective as it could be? First impressions are everything and your landing page serves as that for your business. The landing page is the first thing that your potential customers see when they visit your site. It is critical to grab their attention fast or they will leave your site. On average, the bounce rate, or the percentage of visitors to a particular website who navigate away from the site after viewing only one page, is between 35 to 40%. Ideally, the best bounce rate is below 25% but fewer than 6% of websites achieve that rate.  Aside from bounce rate, the other thing to consider is once people decide to stay, are they actually converting and engaging with your page.  Chances are your website isn’t reaching its full potential.  The best path to increasing performance is through regular A/B testing.
 
A/B testing is the process of running a simultaneous experiment between two landing pages or two elements of a webpage to determine which performs better. The test is named for the two website versions that you are considering when running your test. A/B testing your website’s landing page is one of the best ways to decrease your bounce rate and will also give you the opportunity to test which Call-to-actions and other content resonate best with your site’s visitors.  The results of these tests can greatly impact on your business.
 
A/B testing your landing page isn’t as scary and work intensive as you might think. Here are 5 structured steps to performing an effective A/B test according to ConversionXL:

ABTestCycle

 
1. Find Your Pain Points
 
It’s important to understand what is happening and why it is happening in terms of your landing page’s performance. There are many tools available that you can use to learn about how users are interacting with your webpage. Google Analytics, Mixpanel, and Heatmap are just a few. We’ll be covering those in more detail on the Vlog this week. Once you start digging into the data, you’ll want to look for pages that have a high drop off rate or low conversions. You’ll need to clearly define problem areas that are causing your visitors to leave your site or not convert. Some reasons why conversions might be dropping off include:
 

  • Form fields are too complex and ask for too much information. Less is more.
  • There are too many steps to complete a process. The shorter the better.
  • Page length can be an issue if your content is not simple and clear.

 
Identifying the pain points can help you develop your goals for running an A/B Test. Your goals can be anything from generating a sale to collecting an email address to simply clicking a button.
 
2. Form a Hypothesis
 
Now that you know what you want to accomplish, you can start brainstorming ways to achieve the goals. The first step when performing an A/B test is to form a clear hypothesis. Your hypothesis should stem from your problem that was defined above. A good hypothesis is testable, measurable, has an objective of solving conversion problems and can gain market insights. An example hypothesis could be, “Including more images and photographs on our landing page will lead to higher engagement with prospective customers.” Come up with a list of hypotheses to test and rank them in order of what you expect will happen and how easy it will be to implement.
 
3. Create Landing Page Variations
 
abvariations
 
Now it’s time to put your hypothesis to work and determine specific portions of your landing page to A/B test.  Some of the best elements to test are:
 

  • Call-To-Actions
  • Copywriting
  • Forms
  • Layout
  • Product Pricing
  • Images
  • Amounts of content on the page

 
There are a lot of tools that now make it easy to run these tests such as optimizely.com, Landerapp.com, or VWO.com. Most of these tools have visual editors to make swapping out elements super easy. Also make sure that you are testing mobile and display landing pages separately as users behave differently in each environment.

 

4. Launch Your Test
 
When you’re ready, set the test live and watch how users interact with the variations for a specific time period. Each time a user arrives to your website, 50% will be served your current landing page (or control), and 50% will be served the variation page.

 conversionab

 

Now be patient and let users engage. Don’t be too quick to end the test early based on initial learnings. Wait until the test is over to make decisions and implement changes.

 

5. Measure Results
 
Once your test is complete, compare the results of the control page to the variation page.  If your hypothesis was right, start implementing those changes. You can also use those learnings to find similar opportunities on other pages of your website.  Now that you know what works, you can also try testing out different iterations to further increase results. If the test didn’t work, you can still use that information when forming future ideas to test out.

 

5. The Golden Rule: Always Be Testing!
 
Last but not least, A/B testing is not a one-time deal but a continuous cycle. The effectiveness of a result can change over time and there is always room for improvement. By continually testing, you will identify new opportunities and ultimately increase your bottom line.

 

What A/B Tests have you run and seen the most success? Comment below!

 

Sources:

Optimizely.com

Kissmetrics.com
Unbounce.com
SearchEngineWatch.com

 

Comments