Originally published on Condorly by Anna Jenkins

So you’ve decided to implement conversion rate optimization so you can maximize the amount of visitors who will convert into profit! You’ll probably want to run some A/B tests, increase site speed, and reduce cognitive load. A/B testing is a great option because it allows the tester to change variables one at a time, and makes tracking success pretty easy. But sometimes endless A/B testing won’t solve your problems, so what’s next?

Multivariate testing is usually an outlying option because of time constraints, and is typically harder to deduce the reasons why it worked better than the control page. But when you’re faced with either extremely low traffic, or conversion rates, it’s a good time to head back to the drawing board and start multivariate testing.

Read more to find out about some incredible multivariate test success stories:



Here are 5 case studies that show incredible growth through multivariate testing:


Clarks, a shoe manufacturing company, used multivariate testing in 2010 in order to create a customer experience optimization strategy. Their goal was to improve the conversion rate of website visitors by enhancing the customer experience.

They completed around 36 major multivariate tests; the most recent of which had 160 variables!

As a result, they achieved a 4.2% increase in conversion rate and a 900% improvement in newsletter signups. (Which would naturally lead to even more paying customers).


Nitro ended up applying multivariate tests to their checkout process, which allowed them to see if components such as price points, cart layouts, and cross-selling had a significant impact on conversions.

Over the last two years and after other multivariate tests, they experienced an 88% improvement in in average daily conversion rates and a 380% increase in average sessions per day for each version of Nitro Pro!

Provident Hotels and Resorts

Their goal was to see which combination would instigate the most visitors to check for rates and availability of rooms.

They tracked this test by monitoring the clicks on the CTA button, and ultimately produced 12 combinations to be compared against each other.

As a result, the final combination recorded a 9.1% improvement in conversion rate!




W3i.com (now NativeX), a software company that increases revenue, distribution, and engagement for Windows applications and plugins,used to multivariate testing to refine the performance of their page.

W3i.com wanted to increase the conversion rate on Pimp-Profile.net, a website that featured one of their free desktop applications.

Based on analysis of their previous data regarding their landing page, test results, web analytics, and pay-per-click data, they identified changes that should be made.

As a result, they produced 44 test page variations, 43 of which proved to be a significant improvement from the original page. The ultimate page improved conversion by 162%!


Before (Source)


After (Source)


Hyundai used multivariate testing to optimize the conversion rate on their lead-generation pages.

Hyundai’s car landing pages have a lot of different elements, so it was best to implement multivariate testing. This enables conversion rate experts to observe the way the elements work together when changed.

Their goal was to increase clicks on their brochure request, clicks on the request for a test drive, and a click-through from the car page to the first step of the funnel. They also checked bounce rate as an indicator of engagement.

They ultimately created eight different landing pages to compare against each other.

The result of this multivariate test was incredible success. The request for a brochure/test drive increased 62%, and the click-through rate improved by 208%.


Before (Source)


After (Source)


  • Multivariate testing is a good way to compare different versions of a page in which multiple elements are changing
  • When implemented with a solid game plan, multivariate testing can yield incredible conversion growth
  • When there are many separate elements on a page, multivariate testing is a good option in regards to conversion rate optimization tests

Thanks for reading.