Do you want higher conversion rates from your landing pages? Is your website not generating the revenue you expected? Before you scrap everything and start from scratch, which can be expensive, simple A/B testing can provide valuable information on small changes you could make to achieve more desirable results.
Basics of A/B testing
A/B testing is measuring two variations of a campaign to see which one works better. Campaignmonitor.com did an A/B test on two different email surveys and the only thing they changed were the words on the Call-to-Action (CTA) link found at the bottom of the email. They discovered that one produced 50 percent more conversions than the other. This is an example of how little changes can go a long way. However, if you’ve never tested your site, the first things that should be examined are items that will yield the largest impact. Things like your value proposition, headline and content have the potential to give the highest incremental improvement to your campaign.
The Dos of A/B testing
When performing an A/B test, it is usually in your best interest to only test one variable at a time. If you try testing too many things at once it may be difficult to determine which change produced the results. For example, changing the font on your landing page and adding a background image. Conversions rose 20 percent so you feel as though both changes were for the best. What if the font you chose actually deterred people from the site and your conversion rate would have rose 50 percent if you had just added the background image?
Brainstorming. The first thing you need to figure out is what you want to test. Do a little research and see what other sites are testing and what their results are. Check out your site on Google Analytics and see how people are reacting to your site. If you are new to A/B testing then start with something that has the biggest impact on your viewers. Typically, this is a value proposition or a headline. By improving these you will receive some of the largest return on investments (ROI).
In this phase you should determine how long you want the test to run. You want it to run long enough to gather a good sample size. So for a niche product, this could take a couple of months, where as a mainstream brand may only need a week.
Create/Run. The next step is to create your test page and set it into action. Make both versions available on the Internet and have people randomly directed to one or the other. You can follow the results with a free tool like Google Content Experiments, or you can use a program that has more to offer like Optimizely. Be sure to run the test as long as you initially planned. Don’t stop the test early if you see a trend forming early, as things can change over time.
Analyze and Promote. Once you finish the test, check out your results. If you didn’t see anything significant maybe you need to run the test for longer or maybe the change was insignificant. You may need to revisit your decision and try something else. If you did receive some significant findings then promote whichever page had the most desirable results.
Start all over. Once you complete one test then move on to the next. Start the process over and find something new to test. A/B testing is relatively easy and inexpensive and can provide tremendous ROI. Keep testing!
Cautions of Testing
Google has their bots or spiders that crawl through the Internet to find new or updated pages to keep their rankings current. Since you are testing your page for conversions these bots will have no relevance to your test and there is a way to block bots or have them directed to a specific page so as to not interfere with your test. Do NOT do this! Never try to block bots or hide your page from them. These are necessary for search rankings and if Google thinks you are manipulating it you will be severely punished in the rankings or even removed.
If you want Google to ignore your test page then add content=“noindex” to your meta tag and Google will drop it from any search results. If you are testing a whole website then use the <link rel=”canonical” href= “control page”> between your header tags. This will keep Google happy by telling them which site is the dominant one.
In the end, always keep testing. There is no such thing as a perfect website although there are many very good websites that are successful through testing. For help with A/B testing, contact the professionals at esd & associates who will create and execute a plan to bring your website up to the volume it deserves.
Written by Jerrime Castle
Recently working as a free-lance journalist covering video game news and reviews, Jerrime Castle is currently interning on esd’s Content Development team. With degrees from the University of Washington in Creative Writing and ITT Technical Institute in Visual Communications, Jerrime possess a wide range of both creative and technical knowledge.
