There are many factors involved in creating an efficient website and deciding which of several options will be more favorable requires live testing. A common direct mail method, A/B Testing, can easily be applied to web pages as well.
The current page is used as the control (A) sample and multiple tests are run on test (B) pages to see the results of various changes. In this way, it is possible to test different layouts, color schemes or graphic elements to see which may improve performance.
While there are many A/B Testing tools available, Google supplies a free application called. As with all things Google, it is user-friendly and provides step-by-step instructions. If you already have a Google ID, you can use it to sign into the optimization site, otherwise you will be prompted to create a new account.
Points To Ponder:
Whether you decide to use Google’s optimizer or some other tool, pre-planning is essential:
1. Choose which page you want to test. It needs to offer the visitor some action that will cause them to click through to a second page.
2. Choose your conversion page. This is where the visitor arrives after clicking the link, demonstrating that your test page is successful.
3. Create alternate versions of the control page to be tested that lead to the same conversion page. The website optimizer will track each version to see which one is more likely to have the desired effect.
4. Google’s Optimizer allows you to limit the percentage of visitors who see the new version and most other tools will do the same. However, if your test page traffic is too limited, it will take longer to get meaningful results.
Steps To Follow:
Once you have your pages ready to test, if you are using Google’s Website Optimizer, you sign in and follow these steps:
1. Create a new experiment.
2. Enter the URL of your original page as the first test page.
3. Enter the URL for your new test page.
4. Then enter the URL of the conversion page.
A/B testing has the advantage of being simple to implement with a great deal of flexibility in variable definition. The disadvantage is the necessity of testing only one variable at a time, making it difficult to judge how various elements may interact.