Taking Advantage of Google’s Website Optimizer

As an SEO, one of the main objectives of my job is to drive traffic to websites.  Content, linkbuilding, and traditional SEO are all among the techniques you can use to help you increase traffic and sales.  At the same time, you can also use conversion rate optimization to help you increase each visitor’s value and get the most bang for your buck.

Conversion rate optimization focuses on making your website more effective at getting visitors to perform desired actions.  Whether you want a visitor to download an eBook, sign up for a newsletter, purchase a product, or just move on to another page, conversion rate optimization puts buttons, text, and images in the right place, at the right size, and in the right color to induce a greater number of conversions.

While there are some fairly reliable rules for conversion rate optimization (use white space wisely, make your button contrast with the rest of your website colors and jump out) what exactly works for your site will vary.  After all, each site has its own layout and objectives.  Fortunately, you can figure out what works by testing your tweaks and layouts against one another as part of A/B or multivariate testing.  Google’s recent incorporation of their free testing product, Google Website Optimizer, into Google Analytics has made this easier than ever.

To use Website Optimizer (now called Content Experiments) log in to Analytics and make sure you are on the Standard Reporting dashboard.  Click on Content and you’ll notice a new tab for Experiments.

This is where you find Google's Content Experiments in Analytics

To set up your Experiment, you’ll have to create Goals in Analytics (or use existing ones).  If you want to increase the number of visitors who submit a Contact form, for example, you will need a Goal that tracks how many visitors get to the Thank You page that appears when you submit your Contact form.  If this is already set up, you can move on, if not, establish your Goal (visit here for official instructions from Google on Goal setup).

Now let’s say you want to figure out if a layout of your homepage with a different “Contact Us” button performs better than your current homepage.  Testing this out with Content Experiments is easy.

First, create a page on your website with the new layout that you would like to test and name it (let’s call it yoursite.com/test.html).  Once you have the URL, you plug it and the control page (the version of the homepage that is currently live on your site) into Content Experiments.Choose experiment type in Website Optimizer

As you can see, if you have other variations you’d like to test, you can.  Plug in up to 5 variations of the page you are testing against.

Once this is done, you select the Goal you are using for the experiment (in this example, how many people get to a contact form’s Thank You page), and how many of your Visitors you want included (50%, 100%, etc.).  At this point, depending on your familiarity with coding, you may need to contact your developer, hosting company, or SEO company (hi!) to install some tracking code on your site.  With new Content Experiments, the amount of code required for installation is extremely small, and Google will auto-generate it for you.

Once your code is setup and validated, your Experiment is ready to go!  Visitors participating in your experiment will be randomly served either your original page or one of the variations you provided.  Google will then track which of the variations is the most successful at meeting your Goal.  One of the best things about having Content Experiments incorporate into Analytics is that checking in on the Experiment and generating Reports is especially simple.  You’ll be able to see the experiment’s progress as it is running and get a sense as to which page is winning out.Example report from Google's Content Experiment

(Image taken from Google’s Content Experiment setup instructions https://support.google.com/analytics/bin/answer.py?hl=en&answer=1745152)

Once the experiment is over, you’ll be able to make a well-informed decision about which version of your homepage is the best.  The winner might surprise you; often our personal design preferences are not necessarily those of others, and in many cases a site owner can just be too close to their site to see where they are losing visitors.  It may be obvious to you how to download a newsletter or get to the Contact page, but your users may be experiencing some difficulties.  With Content Experiments you can learn a lot about your visitors and improve your site’s usability.  And now with Context Experiments integrated into Analytics, it’s never been easier to set up an Experiment or access the data.  What’s not to like?

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply