How Google's 'Content Experiments' Can Simplify Website Testing

Google Analytics has gotten a little more, well, analytical. The search giant recently unveiled Content Experiments, a new feature that works with a business' existing Google Analytics account. It allows users to experiment with page layouts to find those that best accomplish online goals such as completing a sale, downloading a file or signing on to a newsletter.

Experienced Google Analytics users should find the Experiments feature familiar. It's based on tools found in packages such as Google Website Optimizer, which is scheduled to be taken offline later this summer. Experiments is being rolled out now and Google says all users should have access to this free tool in the next few weeks.

Here's a rundown of how business owners can start putting this tool to use for their websites: 

Set up an experiment.
Google has placed Experiments under the "content" tab of the Google Analytics account, and serves up a simple four-step process in creating a test. All a business owner needs to do is input the Web address of the page to be tested, and then the variant pages. Simple names such as shop.com/store, /store1, or /store2 work fine. 

Content Experiments can systematically test different page designs against each other to see which drive the most pre-defined conversions. Once the pages to be tested are selected and the goals in Analytics defined, Google provides a few lines of code to be added to the tested pages. Google then begins randomly displaying the variant pages and collecting data on which page designs are the most effective.

Businesses determine how much of their site's normal traffic -- anywhere from 10 percent to all of their regular visitors -- they want directed to the various test pages. Content Experiments then generates comparative results in graphical and data format.

Use the data to create a more efficient website.
This new tool can display different variations of essentially any Web element -- say a photo or bit of copy -- and compare results. The more successful a page design is, the more that page gets displayed, leading to more efficient conversion rate to a business' goals.

For example, you can create a number of different designs for an email newsletter sign up page and then test them against each other to find out which one generates the best results.

Who can benefit most?
Overall, Content Experiments can be useful for testing pages for which a business owner has a clear goal -- such as a sales site or subscription service. Information-oriented sites such as a marketing page, or those with just a few visitors, might not get as much return from Experiments. Having a sizable number of users and clearly-defined goals are required for an effective experiment that tracks user engagement.

Corrections & Amplifications: A previous version of this story misstated the rollout of Google's Content Experiments. Users should have access to the new tool in the next few weeks.