Multivariate Testing

By: Grid14  06-Dec-2011
Keywords: Web Sites

As companies seek new ways to boost online conversion rates and improve their visitors' site experience, interest in multivariate testing is on a feverish rise. But those unfamiliar with the techniques are often unclear about where to start, or how to ensure success.

What is Multivariate testing?

Common methods for running controlled experiments on Web sites range from simple A/B testing to sophisticated multivariate testing, also known as multivariable testing.

In A/B testing, one or more new versions of a page or single site element competes against an existing control version. For example, two versions of a headline might compete against an existing headline.

Multivariate testing, on the other hand, is like running many A/B tests concurrently, where there are multiple elements being tested at the same time. For example, two alternate product images, plus two alternate headlines, plus two alternate product copy text, for a total of 27 possible combinations (including the original control versions).

What's important to understand about multivariate testing is that it not only shows you which combination of elements generate more sales or pull more leads but also reveals which individual elements influence visitor behavioral vs. those that do not. For example, did variations in product image influence visitor behavior more, less, or the same as the copy?

Understanding how each site element causes visitors to interact with your site is the essence of a test-learn-repeat process that marketers can use to synthesize new ideas and continually improve their site's ability to achieve—and exceed—their marketing goals.

Multivariate testing can yield some spectacular results in enhancing online effectiveness.

If you are looking optimize your online performance, multivariate testing should be part of your arsenal of analytics and optimization tools. It may seem like a fantastic revolution in site testing, and it is. There are however some things you should be aware off. First, the sample size required to reach statistical certainty can be large. Second, if you run several multivariate tests across the site at one given time, you exponentially increase the number of variations and the sample size needed for statistical accuracy.

Grid14 helps you to avoid yielding uncertain results and inexplicable outcomes. Should you consider A/B testing or is multivariate the way forward? How to set up a campaign and what are the current software solutions? Our consultants have experience in anything from Google Optimiser to Omniture Test&Target or Memetrics and have the numbers to prove it.

Keywords: Web Sites

Contact Grid14

Email - none provided

Print this page

Other products and services from Grid14


iSchool Seminars

We offer a range of practical seminars to introduce online marketing concepts or to help keep you and your team up-to-date with the latest developments. Online monetization models assign dollar values to desired online behavior, most often those directly involved in driving your overall business. Search engine optimisation is an ongoing process and should form a major part of your marketing strategy.


Grid14 Services

Among other topics, you need to be familiar with web design, human psychology, interaction design, copy writing, statistics, team building, project management and all relevant scientific methods. The biggest problem with Landing Page Optimisation is probably that it does not fit into an organisational chart. It requires a truly diverse set of knowledge and perspectives in order to be effective. Landing pages generally come in 3 main flavours.


Actionable Analytics

It represented the very wise thought that we should not simply stop at the implementation of analytics tools but strive to put extra effort into converting that data into information that is useful for decision making. Oddly enough this step is often skipped or forgotten by most companies, while this is exactly the step that will pay back all efforts and investments in the previous steps.


Landing Page Optimisation

Among other topics, you need to be familiar with web design, human psychology, interaction design, copy writing, statistics, team building, project management and all relevant scientific methods. You should get really excited because now you're getting some idea as to how much opportunity there is to increase your results via constant testing and optimisation.


Behavioural Analytics

Wikipedia defines personas as fictitious characters that are created to represent the different user types within a targeted demographic that might use a site or product. Marketers need to be able to deliver targeted offers to the right customer segments, analyze their effectiveness and quickly make changes based on accurate insight.