Wednesday, March 23, 2011

Which Parts of My Landing Page Should I Test?

With great new, easy-to-use technology like Marketo and Unbounce, marketers can easily execute a key feature of direct marketing: Testing.

But what exactly should marketers be testing about their landing pages? Marketing sherpa Anne Holland shared her wisdom today with webinar participants who checked in to learn about “Winning with Better Landing Pages: Top 5 Secrets to Lifting Conversions.” Here are some pointers from the session.

Button visibility is a major factor in getting visitors to take action, so go to any necessary lengths in size, color, and position to make your “do this” buttons stand out. Simple, less complex landing pages tend to work best (though there are exceptions.. always exceptions), so try testing “distraction removal.” For example, one test Anne showed culled the copy, removed the navigation bar, and pared instructions down to a single, big orange button that said “Get Started.” The result? “Get Started” outperformed its more complicated counterpart by 1,363%. (Yes, one thousand three hundred sixty three percent).

Headline testing is critical for B2B lead-generation landing pages. What seems obvious may be wrong, so marketers shouldn’t rely on their own “best practices” experience. Rather, run an A/B test on headline variations.

In particular, test the headline on your registration form. For example, which header do you think pulled best when featured on a registration form?

  1. Risk Free
  2. Create your own profile for free and unlimited access

Answer: The second.

Marketers often turn the “pretty” part of landing pages over to graphic designers, with the notion that photos and images are simply “window dressing.” Not so. These, too, need to be tested. Should the image be male or female? Show people happy or neutral? Use avatars or real people? Show people or product photos? You can’t know until you test, since all of these (and other) differences affect response.

If you have a forms page, you need to test that, too. Marketers often defer to the IT or database folks about the information that must be on a form. That’s not a good idea, says Holland. “Marketers need to work with the database or IT department on the design of forms. Nothing should be assumed.” For example, should a form have "required fields” indicated or not? One such test showed a 31% lift in forms submitted when the “required fields” notation was left off.

Finally, if any of the following options -- “clear form,” “reset,” or “cancel” -- are offered any where near your form, get rid of them all. This is one thing you don’t need to test, says Holland, who notes that these options (often a hangover from database practices in the 90s) suppress results. “Strip them off,” she says.

Do videos on a landing page suppress or boost response? By now you know, “That depends.” In one case, a ‘form-only’ landing page out-pulled its ‘with-video’ counterpart by 190%. “Don’t just jump on anything because it’s the new thing,” says Holland. “Test.”

For lots more online examples of WHAT to test, check out the

-- scrubbed by Marketing Brillo

No comments: