As communications experts, too often we assume we know our clients’ audiences inside and out. But the reality is that consumer behavior is akin to the weather: it’s influenced by a barrage of external forces and is constantly changing.
In other words, what worked yesterday may not always work tomorrow.
One of the beauties of the digital channel is its ability to adapt to changing behaviors and to test the effectiveness of various Web strategies. Robust testing platforms offer the ability to scrutinize the minutiae and analyze the impact of changes to key variables within the framework of defined goals (e.g. conversions).
Website testing usually takes one of two forms: A/B split or multivariate.
With A/B split testing, a software application splits traffic between two variants of the same web page, measuring whether there is a difference in the effectiveness of that page in achieving the website’s goals. Typically, the test page or element would compete against the original (control) version. Just about every aspect of a website can be split tested.
Some site components you might test include:
- The type of photo or graphic you use on a landing page or other highly valued piece of real estate
- The positioning of a call-to-action button
- The effectiveness of a headline
A different type of testing is sometimes used to achieve more improvements within a shorter amount of time. It’s called multivariate testing. Multivariate testing is like running many A/B tests concurrently, where there are multiple elements being tested at the same time. For example, you could test two product images, plus two headline options, plus two versions of copy about a product, for a total of 27 possible combinations. What’s important to understand about multivariate testing is that it not only shows you which combination of elements generate more sales or pull more leads but also reveals which individual elements influence visitor behavior and which do not.
To maximize a site’s value and effectiveness, you should implement either A/B split or multivariate testing shortly after launch, especially on high traffic pages.
It should be noted that many tests will not be successful against the control. But the key is that with each split test, you are setting a new benchmark. And you’re revealing not only what works and should be implemented but also what doesn’t work and should be avoided.
Once you start testing, you’ll likely find that what you thought would perform well doesn’t. But more importantly, you will be removing much of the guesswork from the equation.
No comments yet
