Table of Contents
The internet has provided huge opportunities for businesses by lowering the barriers to entry and allowing companies to market their products and services to a wider audience. This has also resulted in increased competition.
As a result, businesses are constantly looking for ways to improve the performance of their websites to achieve greater success. Content testing is one of those ways, a way to optimise the performance of your site by adding or removing different elements, altering imagery and messages, font sizes, colours and positioning of key call-to-actions in order to deliver the most favourable results – usually measured by things like lower bounce rates, higher conversion rates, larger basket sizes and lower basket abandonment rates.
Most importantly, testing works IRL. In an award-winning MVT test, Hyundai experienced a 62% increase in conversion rate simply by increasing the size of their images and call-to-action buttons on a page built for booking test drives. Majestic Wine increased their conversions by over 200% on their wedding service page by redesigning it to reduce clutter and testing the new design using a split A/B test.
Split and Multivariate Content Testing
The two common types of content testing are A/B testing (sometimes called split testing) and multivariate testing. A/B tests are a simple form of testing whereby one version of a page (version A) is delivered to a sample of users (typically half the visitors to a site), while another version (version B) is delivered to the rest. After a sufficient number of visits (to allow for the results to be statistically significant), the results of the two experiments are compared to reveal the winner.
Multivariate testing, on the other hand, is like running many A/B tests concurrently. Using multivariate testing, it is possible to test numerous different elements at the same time. So an example of this might be to test a call-to-action button in many different colours, many different images or lots of different messages – often in combination with one another. The same principles apply to multivariate tests whereby the different page versions are compared to identify the version that worked the best in achieving business objectives.
Google Experiments
For sites that have Google Analytics already installed, it makes sense to utilise the in-built tools that Google has developed to allow site owners to run their own content experiments. The key factor in making experiments work for your business is to set up goals on your site.
For e-commerce sites, the obvious goals would be online purchases. However, it is still possible to set up goals for non-transactional websites. For instance, if your receives a large proportion of its income from selling advertising space on a per impression basis, then pages per visit would be a good goal to track. Similarly, if your site relies on offline sales, then customers contacting you through your online forms could be a good way to measure success.
Google Experiments is easy to set up because it uses the existing Google Analytics tracking code to initiate experiments. All you need to do is specify the pages to be used in the test, select the percentage of traffic to assign to the experiment and a minimum time for the experiment to run. More information on Google Experiments can be found on the Google Developers site.
