Summary: This is part three of our four part guide to using Google’s website optimisation tool for better traffic and higher conversions.
Many of the steps used for conducting an A/B experiment are also used to establish the multivariate experiment used within Google’s Website Optimizer.
Step 1: Go to Website Optimizer and Sign-In
To get started, go to Google’s landing page for its Website Optimizer Tool. Again, if you haven’t worked with this tool, the link is: https://www.google.com/analytics/siteopt/splash?hl=en. If you have AdWords, you can also access the Website Optimizer tool there through Reporting>Website Optimizer.
To start a new experiment, you will need to click on the button that says: Create New Experiment and then click on the Multivariate Experiment link.
Step 2: Detail Experiment Pages
Like the A/B test, there are specific pieces of information that are necessary to input in order to identify the experiment pages so that the Website Optimizer tool can do its test:
- Pick experiment name: The name of the first test is not as important as future names, so go for something basic and then add detail with each future test you conduct.
- Submit the URL for the test pages: This is where you will need the URL of the original test page that was selected in the planning stages as explained in part one of this article series. Be sure to leave out other information from the URL like any query parameters as this is not essential.
- Enter the URL of the conversion page: This is the URL of the conversion page that was also selected during the planning stages of this experiment. Also leave out any extra information and just focus on the URL.
At this point, the Website Optimizer will check the URLs for validity. If there are any problems, an error message will be returned and corrections will be made before the experiment can continue. It’s important to get these URLs right to ensure the best possible results from the experiment.
Step 3: Insert Tags on the Experiment Pages
Depending on your level of expertise, you can either choose to install the HTML script tags on the experiment pages or pass this off to someone with more knowledge in this particular area. It’s important to provide all the necessary information, including the URLs, to the person who may be handling this aspect of the Website Optimizer experiment. In fact, there is an actual button where you can pick an option that says: “Your webmaster will install and validate JavaScript tags.” Selecting this option will prompt the Website Optimizer to give you a URL. Along with this URL, you need to give the following information to the person responsible for the tags:
- Elements on the test page that will be varied so that these can be marked with page section tags.
- A name for each element to help identify these elements later on in the test.
There are four types of tags that will need to be installed before the multivariate test can continue:
- Control script: Put this on your test page as a way to ensure that the experiment variations are switched randomly and that these variables are displayed equally. The control script should be placed immediately after the
<head>tag. - Tracking script: Place this tag on your test page to record visits to that page. It should go immediately before the closing
</body>tag. - Page sections script: Use the page sections script to mark the elements that will be varied as part of the test. For each element, the page section must be named in the script.
- Conversion script: Install the conversion script to track visits to the conversion page. Put it immediately before the closing
</body>tags on the conversion page.
Once these tags are completed, Google’s Website Optimizer tool will again validate these actions to verify that the tags have been installed correctly. Any errors will need to be fixed before the test can continue. If everything is correct, you can proceed to the next step.
Step 4: Develop the Variations
The next step in the experiment process is to select the content variations for the experiment. To add a variation, you must do the following:
- Select the “Add new variation” link.
- Type in a name for your variation to differentiate it from the original content and other variations that you plan to test.
- Change the original content that is found in the variation content window to make the variation content for testing.
- Save the variation content.
- Continue these steps until you achieve the number of variations that you want to test.
The number of variations created will also lead to an incremental increase in the total combinations that are analysed within the Website Optimizer tool, testing both the individual variations that were developed and the combined effect of all the variations that had been created. For example, if there were two image variations and two headline variations that were to be tested, then there would be a number of combinations, such as original image and original headline, original image and new headline, original headline and new image, new image and new headline. With even more variations, there are exponentially increasing combinations that can be analysed.
When you are finished with all the variations you want to test, then it’s time to save and continue onto the next phase of the Google Website Optimizer experiment.
Step 5: Preview and Begin the Experiment
Before the experiment begins, you have one last chance to preview each of the potential content and page variations that you have created through the Preview link to make some decisions:
- This link opens a new window where you can see your original test page along with a preview control bar that lets you check out each of the combinations before starting the experiment.
- This is the place where you can limit how many people can participate in this experiment. The Total Traffic Sent through this Experiment prompt allows you to enter a percentage of people who will be allowed to view the page variations.
- This is where you can opt to delete any poorly performing page variations by choosing the Auto-disable losing variations prompt.
- This is where you can hit the Back button and return to earlier steps if the variation pages need to be updated. You may also need to change the tags if any of the changes involve page URL updates.
When you are ready to get the experiment underway, it’s time to hit the Launch Now button. This takes you to the experiment dashboard. From there, you can track how the experiment is going and view a report with data on each of the variations. The amount and quality of the experiment results are based on the amount of traffic that goes to the test variation and conversion pages.
The next article in this series will discuss how to view reports as well as monitor or end a Google Website Optimizer experiment.
Other sections:
- Intro: Using Web Optimiser to increase traffic and optimise content
- Part 1: Planning Your Experiment with Google’s Website Optimiser Tool
- Part 2: Conducting an A/B Experiment with Google’s Website Optimiser
- Part 4: Completing a Google Website Optimiser Experiment and Understand the Results
Enjoyed this article?
Subscribe to our RSS feed, follow us on Twitter or just simply recommend it.
Further Discussion
Leave a Response
Make sure you enter the * required information where indicated. Responses are moderated so please no link dropping, no keywords or domains as names; do not spam, and do not advertise!