Multivariate Testing is Low Hanging Fruit for Marketers
by Eric Dudley on May 15, 2009
Multivariate Testing for Marketers
I was surprised to learn that according to executive surveys conducted by JupiterResearch, only 24% of companies are currently using multivariate testing when evaluating their websites effectiveness.
It certainly seems like low hanging fruit even a 2% improvement in conversion rates could mean millions in incremental revenues. The value of multivariate testing is to yield improved conversion results. In some cases swapping out a photo, message, and/or offer can make a huge difference.
So why are so many companies hesitant to start using multivariate testing tools? According to the survey, the biggest concern from executives is how to recognize ROI with testing. Below is a graph showing the top roadblocks to implementing multivariate testing programs:
Our interactive marketers at WebsiteBiz are experts in multivariate testing. We primarily use Omniture Test & Target (previously Offermatica) and Google Optimizer depending on the complexity of the test and requirements. As marketers have you found similar resistance from executives?
Tagged as:
Google optimizer,
multi-variate testing,
Multivariate Testing,
multivariate testing tool,
Test & Target,
website optimization
Eric Dudley is founder of WebsiteBiz, celebrating 10 years of results-driven Internet marketing. A graduate of University of Delaware, he has led WebsiteBiz to its current leadership position as one of the Southeast’s premier Internet Marketing agencies for performance focused search marketing, email marketing, and online media campaigns. His views on online marketing strategies, search engine marketing, online media buying, permission-based e-mail marketing, website effectiveness, and web analytics have appeared in several publications including Bank Systems & Technology, Charlotte Business Journal and The Charlotte Observer.
twitter
facebook
linkedin
{ 1 trackback }
{ 2 comments… read them below or add one }
I am surprised Jupiter’s number is that high. I’d be interested in knowing what the exact question was. Given the size of the “yes” group I would guess that it allowed a loose definition of “multivariate” or “using.”
Many thanks for this article and for publishing these very interesting stats from JupiterResearch.
I have recently written about this topic in my blog and I tried to identify reasons why multivariate testing hasn’t been adopted more readily here in the UK.
I came up with a few ideas including cost and know-how, but also the issue of who owns the responsibility for sales conversion rather than just traffic to the site. The whole post can be found here:
http://www.culturalmultivariate.com/multivariate/is-multivariate-analysis-arriving/