When used properly, A/B testing can be a life saver for an online business. A/B testing helps you determine the best copy to drive the kind of leads and conversion rates you desire from your websites and marketing efforts.
So, what is A/B testing?
A/B testing: What you need to know
Before defining A/B testing, it is important to first understand what leads and conversion mean to your business.
Online business websites all have a purpose for existence. A blogging site wants to get readers, an e-commerce site wants site visitors to make purchases, a news website wants paid subscriptions, etc.
Before going to into A/B testing, it is, therefore, important that you first define how leads and conversion apply to your business, and then set up A/B tests accordingly.
A/B testing defined
A/B testing is a scientific approach to informing copy production. The test compares two different versions of a web page, app or sales copy to determine which one elicits the desired response from the user or target consumer.
A/B testing is administered by making the two different web/sales copies available to the same general audience, ideally at the same time, and then gauging which one brings the better responses. Brendan at Web Drive Cloud told us, “There are off page factors that can affect conversions, like your web hosting speed, the ad copy that attracted the visitors, and the brand reputation. Getting your landing pages right is crucial though. It’s often about the small changes; the fractions that sway visitors towards action.”
Note: Whatever your testing goal is, remember that you shouldn’t look out for conversion only. Other parameters like click-through rates and time spent on the copy can indicate whether you are making any progress.
Do I need A/B testing?
Yes, you do. It’s hard enough getting traffic to your site. Sometimes, you even pay for it. Whether the traffic is paid or organic, you are investing time, money and other resources to get reached by potential customers. So, if you want to make the most of your investments, it’s time to give A/B testing a shot.
Similarly, A/B testing can help you avoid having your marketing emails ignored or even deleted most of the time. By gauging what kind of copy resonates with your target audience, you can make the transition to being a success story.
In fact, if you need more convincing, then you probably have enough conversions already. Are you ready to A/B test? Then let’s get started.
Where to test
You can perform either onsite or offsite A/B testing.
- Onsite Testing: This includes your web pages. These pages can include your main website, special offers and other promotional copy, product descriptions and landing pages.
- Offsite Testing: This includes your sales emails and advertisements.
Tip 1: Your test subjects depend on the kind of copy you want to test. For example, headline tests are ideal for blog copy, news pages, and promotional copy.
Tip 2: Also, you must choose only test subjects that influence consumer decision for testing. If you are testing on your website for example, it is not a smart move to test just about everything on your website. If you are looking for general guidance on how to build a better website to boost conversion, you can start here.
Tip 3: You must know in clear detail what you hope to achieve from your A/B testing. This means that you should know what results you are already getting from your current copy. If not, you cannot know if a new copy will help you or not.
Choosing your A and B variables
You can use your current copy as variable A and a different copy as variable B. Alternatively, you can come up with two new and different copy as variables A and B.
Note: For each test, there should ideally be only one test subject. If you test more than one thing at a time, it will be difficult to gauge the effects of the changes.
For example, you may test only the call-to-action text in a given A/B test. If you try to test a call-to-action text and a headline in the same A/B test, it can be difficult to know the influence of the variables at each point in time.
Testing more than \ single factor at a time is called multivariate testing. If you are interested in that, then this article is a superb great guide.
Tip 1: You may carry out as many different A/B tests as possible on one test subject before you arrive at a conclusion on which drives the most conversion.
Tip 2: For each test subject, say the call-to-action text, carry out all the A/B tests simultaneously. This serves as a control for the influence of time on consumer behavior. Consumer behavior can be very transient; a time lapse can change a response completely, undermining the authenticity of your tests.
What test subjects to choose
Anything that directly influences consumer behavior. The most important ones include:
- Call-to-action button
- Call-to-action text
- Sales Copy
- Copy text
- Offer contents
Note: Where the audiences are different for the variables, ensure that the two groups are exposed to both variables.
For example, if you are testing a given sales copy with group A and another version of the copy with group B at a given time, then make sure group B is also later tested against the copy you tested group A with and vice versa.
Ideally, you should run each A/B test for only one group at a time, changing only the variables.
For how long should I test?
Depending on how often you run the tests, it can last for several days to a few weeks. The reasons are:
- You want to get data from as many relevant users as possible
- You want to see how much the responses change with time
These make for more accuracy in A/B testing. However, do not carry out A/B tests on a given test subject for too long at a time. The changes in result over time can leave you less informed than when you started.
You can resume A/B tests on a test subject when you notice a drop in conversion, or when you are testing copy for a new campaign or a major restructuring.
A/B testing eliminates guesswork from conversion rate optimization. It enables you to make informed decisions on what works for your business.
Over the next few weeks, we will cover in-depth analysis of the actual A/B testing processes, especially as they pertain to different test subjects such as websites, marketing emails and ads.