Breadcrumb

Multivariate Testing for Ecommerce

Written by

The GoCardless content team comprises a group of subject-matter experts in multiple fields from across GoCardless. The authors and reviewers work in the sales, marketing, legal, and finance departments. All have in-depth knowledge and experience in various aspects of payment scheme technology and the operating rules applicable to each. The team holds expertise in the well-established payment schemes such as UK Direct Debit, the European SEPA scheme, and the US ACH scheme, as well as in schemes operating in Scandinavia, Australia, and New Zealand. See full bio

Last editedOct 20212 min read

When you’re building a website for ecommerce, you’re faced with an overwhelming amount of choice for every element. For example, you need to decide on the best way to organise menus, which photographs to include, how to advertise promotions and even the colours and font styles you should use for icons.

Making these decisions does not depend just on having a good design eye – they can in fact have a significant impact on your conversion rate. That’s where multivariate testing comes in. Using this advanced marketing strategy, you can determine which elements of your website have the greatest impact on sales, and how you can adjust these to optimise conversion rate.

What is multivariate testing in ecommerce?

It’s important to do testing for ecommerce websites to determine how effective they are at generating sales, and to make any changes to improve this as necessary. Essentially, multivariate testing allows you to measure the performance of various different variables at one time to see the effect that this has on conversions. These variables might include:

Checkout or call-to-action buttons. Moving the button to a different location or using a different font style might make it more noticeable.
Within email marketing, you can move the position of headlines, modify images or change text styles.
Within social media, elements such as post length, time and day can be adjusted to observe the effect on performance.

Multivariate testing for ecommerce websites is a complex process that requires vast quantities of data. For this reason, it’s unlikely to work for websites with a relatively small amount of traffic. In addition, sometimes the tests do not identify definite winners. In this case, you can conclude that the element in question does not have a significant impact on conversion rate.

AB testing vs multivariate testing

You might be wondering about AB testing in ecommerce and how this differs from multivariate testing. Put simply, AB testing is a more basic process that may be better suited to smaller businesses with lower amounts of traffic, or those with less marketing expertise. It’s also a great way to generate results fast, although they may be less detailed than those produced in multivariate testing.

For AB testing in ecommerce, two different versions of a website are created and traffic is split between them. The changes tend to be more significant than those within multivariate testing, and so you can observe the differences between two dramatically different websites. Multivariate testing, on the other hand, tends to deal with subtle changes and fine details, and can also study the interactions between different variables.

How to do multivariate testing in ecommerce

There are in fact a number of different tests that may be referred to as multivariate testing, and each has its own unique processes. You can use software to carry out these tests, which will save you having to do complicated calculations yourself. When selecting the right software, it’s a good idea to consider the ease of use, customer support, possible tests, and pricing options.

Full factorial testing is the most common type of multivariate testing for ecommerce websites. Using this methodology, website traffic is split equally among different testing combinations with unique sets of variables. So, if you plan to test 10 combinations, each will receive 10% of the overall website traffic. Since they all get the same amount of traffic, you can easily see which combination produces the best conversion rate.

Partial factorial testing involves exposing only a portion of the combinations to the overall website traffic. The remaining portion is distributed based on the first portion that was tested. This can be a good option for websites with lower traffic.

Using these methodologies, you can take the guesswork out of website design and figure out which approach gets you the best results.

We can help

GoCardless helps you automate payment collection, cutting down on the amount of admin your team needs to deal with when chasing invoices. Find out how GoCardless can help you with ad hoc payments or recurring payments.