The wide availability of setting up A/B tests and multivariate tests has been made even better by tools like Google Optimize and Optimzely. Most anyone can start to run tests on their website and get a great understanding for how to improve their user experience.
While we all want to dive in and test whatever we can, there are a few things to keep in mind. You’ll want to make the testing process impactful to your site and your business.
The following are the top ways we see to fully optimize your testing within Google Optimize. Many of these “rules of testing” span more than just Google Optimize, though some are denoted as Optimize specific.
General A/B Testing Guidelines
Test Creative & Imformative CTAs
The most common element on a site that I see being tested are calls-to-action (CTAs). With good reason too, as the language, size or even color of a CTA can make a large impact on conversion.
With that in mind, make sure you are being creative, informative, and strategic in creating your CTA variants. People don’t respond to the mundane anymore. “Learn More”, “Click Here”, and “Join Today” don’t really tell users what they are doing.
Get better with your testing by calling out the benefits of what your user will receive. You can also talk to them in the first person. Instead of saying “Sign Up Now”, try things like “Access My Account Today”. You never know what will work best with your users, but you will quickly find that the human psyche reacts better when they are treated as individuals and given exactly what they want and expect.
Don’t Test Too Many Variables
If you’ve ever heard the term “Fail Forward” you’ll appreciate my recent experience. I performed a simple test on a page which involved two calls-to-action. These CTAs complimented one another, but at the same time competed. Instead of running a test on one button than the other, or combining them into a multivariate test, I ran an A/B test.
Basically I had four variations of each. The originals were shown together, then variation twos were shown always together, and so on. In the end, the first button’s winning variation was inevitably the second button’s worst performer. I competed against myself in a test.
Avoid this at all cost. In the end I through out the test, it was invalid.
Small Test Can Show Major Improvements
Many organizations hesitate to test minor elements like call-to-action language or button color. From experience, I’ll say that sometimes those are the tests which drive the largest increase in results.
In another recent test, I simply changed the language on a call-to-action button to be more personal and elude to the fact that the sign up was free. In two weeks this new variant was outperforming the original by 34%! Over a year period that equates to nearly 2,000 more conversions for the business.
When you make a decision as to what to test, make sure that complexity is a variable. If something is as simple as changing the wording on a button and can make a large impact, it may be worth the thirty minutes to set it up.
Google Optimize Specific Guidelines
Run the Optimize Script Through Google Tag Manager
Placing the Google Optimize script on your site manually isn’t too tough of task. However, if you want the ultimate ease of implementation, run this script through Google Tag Manager.
Being able to place and execute the script on any page you wish through GTM is a simple process. This also allows you to easily make changes to where you want the script firing, when and how. It gives you more control. If you need help setting this up check out how to integrate Google Optimize through Google Tag Manager.
Take Advantage of Secondary Goals
When you first setup your Optimize test, you will choose a primary goal. This is the goal of what you want your users to do. That may be to signup for a course, or download a PDF, or even purchase a product.
After selecting your initial goal (goals map to your Google Analytics account goals), you can select secondary goals as well. Many people will skip over this as it doesn’t pertain to their main objective.
I would highly encourage you to take advantage of this feature and select a secondary and possibly even tertiary goal. This will help you to evaluate your test on another level once complete.
Say you have a certain variation that doesn’t get the conversion rate on your primary goal you were expecting, but does well on a secondary goal. That can show you what ways the user is being pushed by that variation. This can create a more visible funnel of how you want to progress your users to full conversion.
Secondary goals show just as well in Google Optimize as primary and you can track against the conversions of secondary goals throughout the test.
Use Detail in the Description & Hypothesis Section
Another great, easy to use tool provided by Google Optimize, is right next to the goals and objectives selection. This is simply an open text box that want you to enter a description and hypothesis. Within the tool this does nothing to alter the test itself. Where you’ll love the results of this feature is after the test has run.
Many of us want to continually run A/B tests throughout our site, but we do it just to do it. Utilize this box to ensure you don’t fall into this trap.
Entering a detailed description of the test and what you are hoping to accomplish, accompanied by your hypothesis, will allow for a better evaluation in the end. Try having two soft outcomes within your hypothesis. These can act as a guide in what you will do if one or the other happens. If “X” happens I will do “Y”, if “Z” happens I will do “A”.