Landing Page Testing – Your Key to Higher Conversion
If you aren’t testing landing page variations for your PPC traffic, you could very well be leaving money on the table. Or at least that’s the idea behind landing page testing.
Anyone who has been in the online marketing space for more than a couple months knows that changes to your PPC landing pages can have a serious effect on your conversions – so optimizing your landing pages is key to your affiliate success. But how do you test a landing page?
Andrew Goodman (one of the best known PPC industry veterans) has done a nice write-up about some page testing that he is doing for a client using Google Website Optimizer. He discusses the methodology, assumptions, and elements he used to test 24 variations (YES 24!) of a single landing page simultaneously. While setting up a test might be time consuming, it could be a fantastic way to get your site’s conversions up and raise your profit margins.
Head on over and read the post for a little education and inspiration.
[One note: The Google Website Optimizer isn’t SEO safe (it’s basically a version of cloaking), so don’t do it on a page you are working to optimize.]
| | Permalink | |
|
|
Comments
Hi James,
A great reading tip.
The Website Optimizer looks really great. It is totally independent of Google or even advertising - you can use it with any source of traffic - even with natural search.
The main drawback for me is that (like most other multivariate testing tools) it uses one single web page to contain all the variations, relying on JavaScript to dynamically vary the objects you are testing.
Inserting these code fragments (or chunks, they can be pretty big!) into the web page can be slightly trickier than using two completely separate pages, in particular if you’re already doing a lot of JavaScript on the page.
If you’re testing two or more variations at the same time, it can be difficult to verify that all the variations work and look good together.
Also, when removing the variations you don’t want to keep, you are again required to tamper with the source of one of your most important web pages, which may require additional testing to guarantee that you didn’t break something (or make it ugly in another combination) by inserting and removing critical objects.
For most people. however, Website Optimizer allows you to routinely test variations to headlines, image and button placements, and the formatting of important text segments. It is much faster than simple A/B testing, since you test multiple variations at the same time.
Once you start doing landing page testing, you will easily see increases in conversion rates (and profits) of 10-50%, and you'll wonder why you didn't do this before.
Posted by: Magnus Wester | June 20, 2007 12:03 AM