July 17th, 2008

Of Landing Pages and Google Website Optimizer

At my job, we are constantly looking for ways to improve our landing pages. Our main goal is not to sell something to someone when they arrive, but instead to interest them enough to sign up for our free mailing list. You’d think it would be easy to give something away for free, right?

Well, it’s not! People are very wary of giving their email addresses away, thanks to all those lovely princes of Nigeria who would like to send us money if we would be kind enough to send them a few hundred dollars first. Of course, we never sell or rent the email address of anyone who signs up for our list, but other, less savory companies do. So I don’t blame people for being a bit hesitant before giving up their email address.

The challenge is to convince them that we do, in fact, offer an informative and entertaining service. And that convincing is done almost solely through good marketing copy. It’s been my experience that copy is drastically more important than design when it comes to landing pages, partially because you want to keep your landing pages as simple as possible, to draw the user into the one sole action on the page they can perform: Signing up for your mailing list.

This is when Google Website Optimizer steps in. Once we have a new landing page put together, we’ll run it for a few weeks, as a control test of sorts, to see how it performs. Then we’ll load it into Google Website Optimizer and feed it a few variations. This can be anything from changing headlines around, to using a different image.

Next, Google works its magic by creating the necessary number of combinations based on the different variations you gave it (if I paid more attention in statistics, I could tell you this formula off the top off my head). Then it presents a different combination to each visitor of your page.

Here’s an example of the current test we’re running, with the actual numbers blurred out to protect our marketing director from receiving even more requests for work than he already does:

In this test, we are only trying out different headlines. But you can tell even from the graph how big of a difference a few words can make.

My favorite example of this was a landing page we did a few months ago which referred to the “#1 Stock” We tested this phrase against “Number One Stock”, and found that actually spelling out the phrase increased the conversion rate by 5%!. When you are paying for each and every visitor to your page, that increased conversion rate can mean thousands of dollars saved.

So if you haven’t tried out Google Website Optimizer yet, I urge you to give it a try, even starting with a simple A/B split test to get a feel from it. And if you have used it, I’d love to hear your success stories.

Leave a Reply