Turn More Visitors Into Customers With Multivariate Testing

by blakesanders · 20 comments

This is a guest post by Blake Sanders from Maxymiser.com.

The human brain is a phenomenal thing. It solves problems in ways incomprehensible to most people, and it responds to some information better than others.

Understanding this nature is helpful when website, business, and blog owners look to convert more of their visitors into “customers”. One tool that helps us understand this is multivariate testing.

Unlike A/B or “split” testing, multivariate testing (“MVT”) tests many aspects of a system at a time to learn of the relative impact of that change. The impact of these tests oftentimes leads to higher conversions in the form of newsletter signups, sales, feed subscriptions, and other pivotal elements that increase conversions.

In this article we’ll touch briefly on the impact of multivariate testing.

Multivariate Testing

Who Is Multivariate Testing For?

Multivariate testing can be used by individuals and businesses big and small who want to see how small improvements or adjustments can impact their conversions.

Generally, MVT is best suited towards those who are already generating decent traffic, as this will determine the optimum combination of website elements in the shortest time, giving you significant results to your website’s performance. If you’re not sure the optimum amount of traffic required, multivariate testing calculators are available online and free to use.

Reasons To Test

  1. When you’re fine-tuning: If you want to fine-tune for maximum effectiveness after conducting a/b testing, multivariate testing can work for you.
  2. You want to determine variables that influence each other: If you want to test more than 10 variables at a time, multivariate testing will produce faster results while producing less testing bias.
  3. You’re only able to test specific variables: If you have design restrictions, multivariate testing can be a more efficient strategy than the A/B or “split” testing method.

Split Testing

What Should I Test?

You should first figure the key elements that turn visitors into conversions. Each website’s goals vary, so this varies from business to business.

Are you trying to increase newsletter signups?

Perhaps you want to grow sales?

Say for example you want to test a headline, call to action, and image and the variations of each. Your variations will look like this:

  • Headline:  headline 1 and headline 2
  • Call to action:  call to action 1 and call to action 2
  • Image:  image 1 and image 2

The multivariate tests aims to see which combination of these versions achieves the highest conversion rate. Traffic is split between the following to see which combination produces the highest conversion rates.

8 Testing Variation Combinations -:

  1. Headline 1 + Call to Action 1 + Image 1
  2. Headline 1 + Call to Action 1 + Image 2
  3. Headline 1 + Call to Action 2 + Image 1
  4. Headline 1 + Call to Action 2 + Image 2
  5. Headline 2 + Call to Action 1 + Image 1
  6. Headline 2 + Call to Action 1 + Image 2
  7. Headline 2 + Call to Action 2 + Image 1
  8. Headline 2 + Call to Action 2 + Image 2

Multivariate Testing Tools

To create your multivariate test, first choose a framework that supports it. Afterwards, choose which sections to include in it.

Google Website Optimizer is a free basic multivariate testing tool by Google. It’s the easiest way to see how making changes to page elements has on conversion rates.

Conducting Test: Some Tips

Remembering these simple details can go a long way towards making your test a successful one.

  • Reviewing your data is of the utmost importance. Keep track of reports via spreadsheet and pinpoint the elements that are causing the biggest swing in conversion.
  • Keep up with your competitors. They very well may be testing elements of their own.
  • Be patient. Results come over time. Testing pages that get a lot of traffic and determining the proper length for running a test will give data for measurements.


Multivariate testing is crucial in the evolution of your website to attract new leads. This easy to use tool has the ability to affect businesses revenue and profits. Use multivariate testing on your next online marketing effort to gain insights into customer preferences and user behavior.

About blakesanders

Blake Sanders writes for Maxymiser and is an expert on multivariate testing. Blake enjoys discussing various topics on increasing website conversions.

has written 1 superb article(s) on this blog

Maddy December 7, 2011 at 3:06 pm

Hi. It’s a great Post but how can this help small bloggers ? i mean people who are not selling any thing on their blog. Is it anyhow useful for them ?

Jasmine December 7, 2011 at 3:11 pm

Sounds great. I have not tried any multivariate testing actually because I am not too sure how I should do it. But seems like we can do that with Google Website Optimizer!… free and easy to carry out. Will give it a try later.

Evie December 7, 2011 at 4:28 pm

Excellent, I’m such a massive fan of multivariate testing and anyone who isn’t 100% clued up on it can get a lot of information from this post. Everyone should run some form of multivariate testing to make sure they are getting the most out of their website or blog.

Venus December 8, 2011 at 3:34 pm

I once performed A/B testing for my software but it wasn’t so much successful.Now I am planning to opt for “Google Website Optimizer” after this post.
Thanks for Blake for being helpful.

Becca December 10, 2011 at 11:04 pm

Ah i didn’t even know that they had the tools to do this, but it’s quite good that there are tools out there that save your time. I’ve just found out about this testing and it is interesting how some minor changes can influence the conversion rate or something like that. But I am not sure, for myself I think i will just put the things where i like them to be and leave them there hehe without playing around too much with this..

karen December 13, 2011 at 9:23 am

I will definitely try this one. Visitors are important but we all must admit that we need customers more than anything. Thanks for sharing.

Petter December 13, 2011 at 6:27 pm

Good article! I haven’t tried multivariate testing myself, but I can see the usefulness. I think the big challenge comes when it’s time to analyze your testresults in order to outline a winning strategy.

David December 24, 2011 at 12:20 pm


That’s a really interesting post. I had seen Google Website Optimizer, but didn’t understand the MVT concept until reading this, so thank you very much!

Merry Christmas!


ralph edrozo January 4, 2012 at 10:38 pm

i will try this one , to improve my blog. and also Testing is the best way to learn what works and what doesn’t no matter how sure someone could be about his/her sales copy or landing page. thanks for the idea ..

Noah April 7, 2012 at 12:32 pm

Hey Blake,

Multivariate testing is new to my ears but after reading this article, it really gives me new ideas on how to improve my website’s conversion rates. It is true that patience and reports both plays a crucial role regarding this method.

Thanks for sharing. Worth reading.

Previous post:

Next post: