Increasing Performance and Conversions Through Multivariate Testing

Savvy marketers understand these days it’s not how much traffic you drive to your site, but how many of those visitors are being turned into customers—turning to multivariate testing (MVT) is key in creating conversions.

By conducting multiple tests, marketers are able to determine how well visitors respond to certain elements on their site. If you haven’t jumped on the multivariate testing bandwagon, perhaps now is the time. Simple changes to copy, form layouts, calls to action, and even landing page images can have a dramatic impact on conversions.

Multivariate testing, like A/B testing, allows you to test which element works the best, but has the advantage of letting you test many changes simultaneously. This oftentimes leads to better results that can be implemented into your online marketing efforts.

The Impact of Multivariate Testing

The benefits of using multivariate tests are plentiful; the distinct advantage multivariate testing (MVT) has over A/B testing is that it allows for experimentation of several elements at once. This eliminates a lot of the guess work related to standard A/B tests, where testers change one item at a time and then see if it had any impact on the data.

MVT can also tell you which part of a page is most influential in conversion goals. The multivariate test will give you a metric called an “impact factor” in the report and tells you which sections influence the conversion rate the most. A/B tests simply combine the sections into one variation.

With simple multivariate testing, a website can attain a high conversion rate for: newsletter signups, sales, feed subscriptions, downloads, and many other things. More importantly, it helps keep visitors engaged on a site, enhancing their experience, lowering bounce rate, and helping with website stats and analytics.

Looking at A/B Testing

Traffic for Significant Results

Before you begin testing, it’s important to get an idea of how much traffic you’ll need for the best results. You don’t want to add sections to your page if you’re only getting a miniscule amount of traffic to your site. There are calculators available online to help you determine exactly how much traffic you need for a multivariate test to go smoothly. This is an inherent issue with MVT; having a sufficient amount of users to test particular qualities due to expanding across multiple testing sections.

You should begin testing after your site launches and has a steady stream of traffic. Even if you feel your website is performing well, there’s always incentive (and room) for improvement, and multivariate testing will help you raise the bar.

Design Elements to Test

The elements to test on your website vary and are usually contingent on your website’s goals, although it doesn’t hurt to run tests on all elements. Narrow it down to the few elements essential to your conversion goals; commonly tested elements of a web page include:

  • Headline and heading
  • Call-to-action buttons (color, text, size, placement)
  • Text copy (content, length, size)
  • Image (type, placement, size)
  • Form length

Tools and Starting

The Google Website Optimizer is a free basic multivariate testing tool by Google. It’s great if you want to test the waters before investing money in multivariate testing and includes plenty of video and tutorials to get you started.

The aim of any test is to increase usability for the user and value for a website owner. How this happens may change across testing platforms, generally there will be some standard requirements.

  • Website owners must create the page variations
  • Code must be inserted in each page variation
  • The testing program will display equal amounts of the page variations
  • The program will track the data and produce statistics

Testing will vary greatly across websites, as some websites are designed by hand and others use content management systems (CMS) or other platforms. For more popular CMSs, there are extensions available that work with programs like Google’s Website Optimizer.

Putting Results into Action

This is the easy part. Any MVT is going to indicate which page variation drew the high number of conversions or “goals”. The next step would be to put this layout as the permanent design for your viewers.

These tests will also tell users which elements (headlines, images, colors, etc.) had the biggest impact on conversions. You can even further these results by varying which had the biggest impact to further improve your results.


The window to grab a visitor’s attention has decreased significantly in the past few years. Make sure your online marketing materials are working their hardest so that they can by running multivariate tests to determine the most successful elements. This article is meant to get you started down the path of creating websites and marketing initiatives that convert and get you money. How is multivariate testing working for you? Or do you prefer A/B testing?