- What is SharePoint 2010? Vision and Reality
view comments - Myths & Realities of Drupal
view comments - Knowledge Management in 2012? Probably Dead
view comments - iPad 3 vs. New Samsung Tablet: War Starts in February
view comments - iPad 3 to be Unveiled by Apple in Early March
view comments - 5 Signs Your Company Doesn't Get Social Business
view comments - 5 Critical Steps to SharePoint Information Architecture Planning
view comments - Knowledge Management Meets Social Business: KM is Dead, Long Live KM!
view comments
How To: Getting Started with Google's Website Optimizer
Gone are the days when you needed to create paper prototypes of a page and test them in front of focus groups or run them through usability testing. Technology has evolved this process.
The Google Website Optimizer now helps one test and optimize a website all in the comfort of a real live visitor base.
The Optimizer has been around in beta status for awhile now. We first heard about it back when Clickability put out a PR announcing their addition to the Website Optimizer Technology Partner (WOTEP) program. This WOTEP program has content management providers as partners with Google to help integrate the Website Optimizer with their solutions.
But you can use the Website Optimizer on your own. And here we show you the basics of what the tool does and how it works.
To use the Website Optimizer you need to have a Google AdWords account — so set this up before you get started. [Editor's Note: As of April the Optimizer no longer requires an AdWords account, just a Google Account. The two products are however complementary.]
There are two types of tests you can conduct with the Website Optimizer: A/B Split Testing and Multivariate Testing.
A/B Split Testing
This is the process of testing completely different versions of a particular web page. Google recommends this test if you have less than 1,000 page views a week and/or want to test a different look and feel for your page.
Multivariate Testing
This is the Process of testing different combinations of content in multiple locations of a particular page. Google recommends this test if you have more than 1,000 page views per week or if you want to try different content on a page (i.e., different images, headings, etc.).
Depending on the test you choose, the setup is a little different.

Overview of A/B Split Testing
The setup of this type of test is pretty straightforward. Here are the steps:
- Select the page you want to test
- Create any number of variations of the original page you want to test - each page must have a unique URL
- Add Tags to each of your variations — or experiment pages
- Identify the “goal” page you want to reach - this will inform Google a conversion has taken place
- Run your test
- View Reports
- End your test

Tagging Your Pages
The tagging of your pages requires some knowledge of HTML. There are two scripts to install for this type of test:
- Control Script: To be added directly after the <head> tag on the original page, this script ensures that the experiment pages are switched randomly and are displayed an equal number of times.
- Tracking Script: To be added to the original page, all experiment pages and the conversion page just before the closing </body> tag, this script ensures that all pages are tracked for the experiment
Overview of Multivariate Testing
Setting up the tests for multivariate testing is more complicated than performing A/B Testing. This type of test requires that you add tags to your test page and define a number of potential variations for the sections you have chosen to test. Here are the basic steps for the Multivariate test:
Continue reading this article:
Featured Events View all
| Add event
|
RSS
- Feb 22, 2012 – Intelligent Content Palm Springs 2012
- Feb 26, 2012 – SPTechCon - Sharepoint Conference San Francisco 2012
- Feb 28, 2012 – (Webinar) How to Build Great Mobile Websites
- Mar 6, 2012 – Get Social with Microsoft & Telligent in Dallas
- Mar 8, 2012 – Get Social with Microsoft & Telligent in New York
Who's Hiring? View all
| Post a job
|
RSS
- Web Content Manager in Newport Beach at Orange County Museum of Art
- Principal Business Consultant in Paris at Saba
- Director of Customer Success Management in Nova Scotia at Radian6
- Software Engineer -- Media Solutions in Bucharest at Adobe
- Technical Writer in Charleston at Blackbaud
- Interaction Designer in Maryland at Inmedius
- Project Manager in London at Brandworkz
- Sales Director, Consumer Electronics at Synacor

Receive
the Free CMSWire Newsletter
Email It