In the arcane and technical world of online marketing, techniques to increase customers and therefore profits remain supreme. A website is literally a virtual store, complete with shopping aisles and checkout counters. The website serves the dual purpose of marketing products to visitors and facilitating their purchases.
A website is constantly trying to sell its wares, whether products or services, making it the virtual equivalent of a salesman who works around the clock. Websites must convert visitors into customers in order to maintain profitability. Every website owner pays attention to a key metric in order to do this: the conversion rate.
Conversion Rates And Testing
The conversion rate tells the website owner what percentage of visitors are becoming customers. The rate is the result of dividing the total number of visitors by the number of customers. Conversion rates can spell the difference between profit and loss. Importantly, they can power a website’s sales even if the website has low traffic volume. A high conversion rate will deliver consistent profitability across the unpredictable vagaries of Internet traffic. Improving the conversion rate is a consistent method of increasing profits, no
matter the state of the Internet traffic set-up.
Multiple methods exist of improving the conversion rate, but they all involve testing the website in some way. A/B testing is the simplest and easiest method to use. A/B testing is as simple as it sounds. A single page element is split into two versions and shown to an equal number of visitors. The version that results in the highest conversion rate wins and is implemented on the site proper. A/B testing allows a website owner to get inside the heads of his visitors and discover what they like and do not like. This testing method must be implemented properly in order to work. Plenty of opportunities for error exist, and they can distort the results.
The website is split into two different instances. The first instance is left untouched, but the second has one element changed in some way. Multiple components within a site can be changed. Anything from colors, textual content, headers, footers and images may be changed. Meta-elements, such as the position or configuration of any of the individual elements, may also be altered. The point of A/B testing is to quantitatively figure out what kind of website the visitors like best.
Setting Up The Test
One of the most critical aspects of any A/B test is knowing what factors to test and how to test them. This, in turn, depends on the goal of the website, or what the owner wants his visitors to do. Signing up for a newsletter, purchasing a product or service and completing a survey are all possible goals. These different aims require different testing methods that do not target the same elements. If the goal is to persuade a visitor to purchase a product or service, A/B testing should focus on elements relating to that goal. The above notwithstanding, a few items are usually tested no matter what the goal. Two of these features are call-to-action buttons and website layout.
A/B testing can only focus on one component at a time. There are two ways to implement the test. The item can be changed while the page is loading, or the visitor can be redirected to an alternate page containing the changed item. The first method is useful for testing single page factors, like a color or a button. The second method works best with large-scale changes like layouts. A/B testing, despite its apparent simplicity, is quite sophisticated and can produce robust results.
Never test the control page or the variation page apart from each another. They must be tested simultaneously or the results will be ruined. Since Internet traffic changes so much over the course of a single day, testing the two pages simultaneously is the only way to be certain of the results. Otherwise, a plethora of reasons could be named for the obtained results, making the test worthless.