What a Boomerang Expert Taught Us about Multivariate Testing

Barnaby Ruhe is a fan of freestyling during boomerang competitions and is credited with inventing the foot catch. To accomplish this, Barnaby needed to throw his boomerang and then accurately choose where to lie down on the grass to catch his returning ‘rang between his feet. At the time, this seemed radical because everybody knew that boomerangs were caught by hand, but now this catch is incorporated into international rules. He compares that to the Harlem Globetrotter brand of basketball in which players passed the ball backwards, in between their legs and more. When they developed this style, the focus was largely on creating entertainment. Today, NBA players competitively use these techniques—just as modern-day competitive boomerang athletes use Barnaby’s innovation. Barnaby is a fan of testing multiple aspects of boomerang throwing all at once. This could include changing position slightly to incorporate the power of the wind in a different way while adjusting the hold on the boomerang while also releasing the boomerang differently, perhaps with a coin taped onto a tip of the ‘rang. Each one of these would have an impact on the boomerang’s flight and return, thereby making this an offbeat example of multivariate testing.

Multivariate Conversion Rate Optimization: Website Testing

Barnaby’s foot catch and the Harlem Globetrotter backwards pass both grabbed spectators’ attention—and that’s crucial on the web. People’s attention span is estimated to be six to eight seconds when browsing a web page, with the now-infamous comparison stating that this is even shorter than the attention span of a goldfish. One of the best ways to ensure someone’s attention is appropriately directed on a web page is to conduct conversion optimization testing, with the two most common testing types being A/B and multivariate. When handled well, multivariate testing can be more efficient because, in a sense, you’re combining multiple A/B tests with the same goal, saving you from the necessity of running numerous tests on a page sequentially. You could, for example, test text-based changes and visual changes on a webpage together and/or test the length of a form field while also testing corresponding calls to action. It takes finesse, though; rather than testing one version of a webpage against another with, say, different call to action buttons (as you would with A/B testing), you are testing combinations of variations within the web page. This ultimately allows you to determine what elements on that page in which combinations are most effective in achieving your goals for that page. Barnaby’s experiences suggest that the best results come from a combination of scientific testing with a fresh creative approach that sends preconceived notions out the door. So, first ensure that you are crystal clear about:

• your business goals
• challenges you are facing to meet these goals
• how a specific digital marketing channel fits into a goal
• how much of a sample size is needed to get useful information; remember that with multivariate testing, a larger sample size is typically needed to get solid results
• the length of your testing
• how results will add to knowledge gleaned from previous testing, if any
• how results will potentially fuel future testing

Next, create your hypotheses. Why do you think your conversions are too low? You can determine this in numerous ways, including:

• asking employees to review the site and give honest feedback
• asking other, more objective people to review the site and give honest feedback
• reviewing data from website analytics
• analyzing pages from a competitor that performs well online

You then can begin to formulate tests—and, like Barnaby, we suggest you use a combination of logical changes and ones that are more creative and even unexpected. Pay special attention if someone says that a specific element shouldn’t be changed in a certain way during testing because “everybody knows” that won’t work. Really? Consider testing and, if “everybody” was right, then, great. That hypothesis is now confirmed. If not, then your creativity will likely put you ahead of your competitors, and—who knows?—your creative testing today may just lead to creating tomorrow’s best practices.

Conversion Rate Optimization Services

If you need help with your conversion rate optimization, here’s more info or contact us with questions! We’re here to help.

The post What a Boomerang Expert Taught Us about Multivariate Testing appeared first on DAGMAR.