Share on LinkedIn0Tweet about this on TwitterShare on Facebook0Google+0

Multivariate Testing FAQs

multivariate-testing-01

If you find the prospect of conducting multivariate testing on your landing pages a little daunting, we understand. After investing hours upon hours into writing, editing, and finalizing a piece of content, you’re probably pretty anxious to push it live. Add on the tasks of uploading it, creating a form, and building a landing page…well, the last thing you want to do is add another step to the process.

But consider this infuriating possibility: after all your hours of hard work, the tiniest flaw in the wording or design on your landing page could be keeping keeping prospects from converting, meaning they’ll never see the content that you worked so hard on. Multivariate testing on your landing pages can help you reap the full benefits of your content — and that’s worth the additional time investment. Let’s a take a brief look at some frequently asked questions to help you get started.

How does it work?

Simple. Multivariate testing helps you develop the best possible version of your landing page by comparing the effectiveness of different variables. You create several slightly different versions of the same landing page, and your marketing automation platform randomly assigns different versions to different visitors, tracking each version’s success.

How is it different from A/B testing?

A/B testing compares two versions of a single variable to show which version is more effective, while multivariate testing compares multiple variables in a variety of combinations. Consider the following examples:

  • A/B Testing: Do more visitors sign up for a free demo or a white paper?

  • Multivariate Testing: Do visitors respond better to a short form next to an image of the white paper on offer? A screenshot next to a long form that earns them a free demo? Which of these options fares better once the headline has been worded differently?

Although multivariate testing may take longer to set up, results tend to be more thorough and revealing.

What should I test?

There are multiple aspects of a landing that can— and should — be tested (images, form length, copy, value proposition, etc.), but consider starting off with the following three variables:

  • Layout: Research shows that your audience processes web pages in the pattern of a capital “F,” meaning that your most important, attention-grabbing material should be focused in the top left-hand side of your landing page. That being said, what’s your most attention-grabbing material? Move things around and test a few different layouts to find out.

  • Headline: As we noted in our post, “6 Tips for Effective Landing Page Copy,” your headline should always match the copy on the link or advertisement that led a prospect to your landing page so that your visitors don’t feel misled. But don’t feel limited by this rule: even the slightest difference in wording or phrasing can make a difference, so test a few variations to find the most compelling headline possible.

  • Call to action button: It may seem like a tiny detail (it’s probably only two or three words!) but the phrasing of your call to action makes a big difference, and since this is the last step before conversion, you’ll want to get it right. Check out a few of our resources on writing effective CTAs and draft up a couple versions to test.

Will my visitors notice?

No. Visitors to your landing pages are cookied so that each time they return to the landing page, they will see the same version of the landing page they saw previously.

How long will testing take?

There’s no set time for running a multivariate test, you just want to make sure that you have a large enough sample size to secure accurate results — and that’s a matter of personal discretion. Luckily, once you’ve set everything up, it’s just a matter of checking back on your reporting periodically to see how each page is doing.

What’s being tracked?

Reporting will vary from system to system, but in general, you’re looking for statistics on views, conversions, and conversion rates (Pardot also combines these factors to automatically determine the top-performing landing page). It’s important to pay particular attention to conversion rates in order to pinpoint the factors that are encouraging visitors to convert.

How often should I test?

Continuously. Like most aspects of marketing, landing page success relies on agility — you should be constantly reevaluating and adjusting. Are there different word combinations to try? New items to introduce? Layouts you haven’t yet experimented with?

Sound exhausting? It won’t be. After you’ve set up your initial multivariate test, experimenting with new variations is simple — and interesting! Seeing the difference that a seemingly insignificant word choice makes can be fascinating, and can reveal a great deal of insight about your target audience.

What are your experiences with multivariate testing? We’d love to hear your thoughts in our comments section.