This article was written by Rebecca Q Jones from the A/B and multivariate testing website Maxymiser, enabling site owners to increase the effectiveness of websites by using multivariate testing.
A/B and Multivariate Split Testing – What Are They and Why Are They Important?
A physical store offers goods and services to visitors, hoping to convert them into customers. The Internet has transferred this function to a virtual store, a website that sells goods and services by way of an Internet connection. Potential customers can access the virtual store more quickly, and they can be converted to customers more easily.
Every online business is focused on a key metric that measures quantitatively how well this goal has been met. The metric is known as the conversion rate, obtained by dividing the number of customers by the number of visitors. The conversion rate can be a website’s key to business success.
Even with a small traffic volume, a high conversion rate can still translate into greater profits. To improve the conversion rate, it is essential to understand the rate as a psychological effect. The website has to persuade a visitor to become a customer by making a purchase, joining an email list or signing up for a newsletter. Every element of a website works together to persuade the visitor. Persuading every visitor in just the right way is obviously impossible.
Business owners can conduct tests to discover which elements in what configurations and settings have the greatest impact.
A/B Testing
This method is straightforward: A single element is split into two versions and both exposed to the same number of visitors. The version that converts the highest number of visitors wins. A/B testing is simple to implement because the website tests one element at a time. The A/B method can handle a specific element on a page or a page-wide element like color or theme. When testing a single element, replacing the element to be tested before the page loads works best. When testing a page-wide element, redirecting to another page is the way to go.
Multivariate Testing
Where A/B handles the building blocks of a web page, multivariate testing takes this to a new level. Multivariate testing takes multiple elements and groups them into combinations. This method tests all of them simultaneously to see which combination of elements yields the most conversions. Multivariate testing is much more complex than A/B testing for obvious reasons. A multivariate test needs a specific, defined goal in order to succeed. A number of factors can influence the results of a multivariate test. These factors include the number of visitors, the existing conversion rate and how long the test has been running.
Analyzing The Results
Both tests use statistical analysis to analyze the results. Using common sense to learn from the results also helps. Statistics cannot reveal everything a business owner needs to know. Disparate elements, including images, colors, content size and placement all play a role in persuading a visitor to become a customer. A/B and multivariate testing help a business owner meet this goal and improve the conversion rate.
Options
Services like Google Website Optimizer and Visual Website Optimizer can help business owners test their websites. Website optimization is a new trend in online business or e-commerce. Conducting business online is much simpler than trying to run a physical store. True, the concerns are the same, including suppliers and customer service, but a website is still different.
Google Web Optimizer
Conclusion
A/B and multivariate testing take online business to the next level. Using these testing methods breathes new life into a website and into the business. Online business websites and physical stores share one thing in common: Making the customer happy. A/B and multivariate testing use precise methodologies to meet this goal and increase the website’s profits. In the final analysis, these testing methods update and replace older market research techniques. You can learn more about Multivariate testing over at the Maxymiser multivariate testing guide.
Trackback URL for this post: http://www.bizchickblogs.com/2011/07/split-testing-your-website-an-introduction.html/trackback/