Abandon basket campaigns are a key retargeting...
It goes without saying that your website is a key touchpoint in the customer lifecycle - and requires significant contribution from your budget. Research suggests that the time you have to capture the attention of your customer on your website is less than 9 seconds. That means you only have 9 seconds to engage your customer to ultimately carry out the goal conversion you have set. It will always be a priority to test the ability of each element of your website to grab the customers constantly sought-after attention.
Naturally, usability testing maybe your first point of call. This involves conducting qualitative experiments with a sample of users, observing them guide themselves through the site, and completing the tasks set out for them. However, a quantitative method will help develop a more convincing ROI model. Further still, a combination of these testing methods will help you build a robust strategy for making truly behaviour-based additions to your website. This is where A/B and multivariate testing comes into play. Each method comes with its own merits and limitations, so we’ve drafted in our resident data expert to help us clearly understand the reliability and usability of each.
We’ve asked our Data Scientist, Phil, to help explain how MVT can up your customer engagement game.
How do you know your website is performing to its highest conversion ability?
If you want your website to continually engage your customers, you need to test which elements of each page will achieve your objectives and drive conversions. You will also need a test that will provide a sharp and accurate measurement for determining ROI. In this instance it would be wise to use something like a multivariate or an A/B test.
What is multivariate testing?
Multivariate testing (or MVT) is a method for understanding the interactions between multiple variables on a page in order to find the best overall combination that will satisfy your site objectives. Unlike an A/B test, a multivariate test means you can test two or more elements on a page at the same time. This allows you to see which combination of variables drive the best outcomes for a particular goal on your site. A multivariate test can help you get the results that you could in part get from potentially conducting multiple A/B tests over time.
You may want to test, for example, the impact of two different headings, two different images, and two different button placements. You would have to conduct 3 different A/B tests to discover which version of each gets the best result. Conducting a multivariate test will help you learn the best combination of variables that will engage and convert your customers.
If MVT is so comprehensive, why do we bother with A/B testing?
There are some limitations to multivariate testing. As you are testing lots of different variations on a page, the sample size is greatly reduced, as traffic needs to be equally divided. Therefore, in order for your results to be statistically significant, a high volume of site visitors are required. Understanding impact can prove even more difficult once you need to drill down into less frequented pages. Both of these issues mean that a lower statistical significance will be reported – so results could be unreliable. Finally, as you are working with a large number of variables, conducting a multivariate test is a more complex process than A/B testing and may take up some additional resource. It’s important to have a full understanding of your conversion goals before you commit to testing in-house.
Okay, so what type of speciality resources do you need to do this?
One method will not be a catch all for every scenario, it’s advisable to access a team that is confident in conducting testing using various web analytics and usability tools. Quantitative testing is imperative if you want to provide fool-proof evidence of your digital investment, ensuring future budget allocation to your team.
Intilery will easily integrate into your current platform so you can create and deploy A/B, multivariate testing, with intuitive reporting features. The technology behind Intilery is powerful, agile and adaptive, and the team also offer expert advice through its data Science function. Get more information here.