Multivariate Testing Accelerates and Drives Improved Conversion Rates
As online marketers we have many levers to optimize. As we strive to attract qualified visitors we also realize a targeted visitor is only as good as the experience they encounter when they arrive. It is our opportunity to deliver an exceptional experience and great offer that they choose to act and achieve a desired outcome.
Testing landing pages is a great way to determine what message, experience, and offer our audience best responds. It's about "lift" and "improved conversion rates". By thoroughly testing and optimizing landing pages, marketers have realized exponential returns.
While an effective, but less sophisticated model of testing, A/B testing tests a single site element against a single control version. For example, two landing pages each with a different headline and hero image.
[caption id="attachment_119" align="alignright" width="300" caption="Quickbooks Multivariate Test"]
[/caption]
[/caption]
Multivariate testing takes A/B testing to another level by testing multiple elements at the same time. It not only shows you which combination of elements generates more leads, but shows which individual elements generates more conversions, whether it be a call to action, a button, an offer, or even a hero image. Multivariate testing allows you to identify the specific aspects of your campaign elements that have the greatest influence on site visitors.
By using multivariate testing, you will be able to ascertain which elements of your Paid Search landing page yields the highest number of site visitors to perform a desired action such as requesting information or making a purchase.While many of the champion combinations may seem intuitive, many will also surprise you.Using multivariate testing accelerates a marketers ability to yield the best results when comparing to A/B testing. As an example, with one of our clients, we were able to produce a lift of over 11% using multivariate testing in a short period of time. If you are interested in learning more about multivariate testing we would recommend these:
- Google Optimizer is a free multivariate testing platform but has many limitations.
- Omniture recently purchased Offermatica and have re-branded it: Test & Target. It's a much more sophisticated platform for multivariate testing and one we have evaluated and adopted.
Comments
“Using multivariate testing accelerates a marketers ability to yield the best results when comparing to A/B testing.”
Eric- I appreciate your adoption of MVT. I’ve spent much of the past three years evangelizing it myself. However, I wanted to point out that the statement above is not necessarily true. In fact in most scenarios the opposite is true. Many times A/B testing can accomplish better results, with higher confidence, in much less time.
As an example, the Quickbooks tests you reference as MVT were actually A/B/C tests.
Happy testing!
Hi Eric, nice post - very informative!
Yes, a 1% increase in conversion rates is very good, but a 11% increase is quite significant-Congratulations! We’d love for you to take a peek at SiteSpect for A/B and multivariate testing (check out: www.sitespect.com).
We do it without page tagging, so tests are run in a matter of minutes without IT involvement; and it’s a sophisticated platform run as either SaaS or as an enterprise server. Anyway, just wanted to drop a comment to let you know that website testing and targeting can be done quickly without the page tagging.
Thanks!
Jason O’Keefe
Jonathan- thanks for your comment on multivariate testing. I agree that the Quickbooks example was not the best for illustrating multivariate testing.
That is not easy to get the highest page ranks. You would not be able to do that if you are not familiar with search engine optimization. Do not trouble about these things! You can find professional forum posting service to increase traffic to your website. Experts realize a way to do it. Therefore, there’s not nothing to worry about!