Perform Content Experiments in Google Analytics

In early June, Google sent an email around to users announcing that you can now create A/B/N experiments (“Experiments” or “Content Experiments”) to measure the effectiveness of different versions of your web pages! This new feature is incredibly cool and, unlike traditional A/B testing or multivariate testing, it allows you to test up to five full versions of a single page.

Why is this feature so awesome?

Using Google’s new Content Experiments feature, you can compare how different web pages perform and optimize your site to increase conversions. For example, let’s say you’re a local plumber interested in improving your website’s visit-to-appointment conversion rate. Typically, you use a coupon to encourage your visitors to schedule their plumbing appointment online. With Google Content Experiments, you might create one page with mega-sized coupons, one with the coupons in a different location on the page, and one with an additional call to action above the coupons.

With your tests in place, you will now select what percentage of your web visitors are included in the experiment. Google will then show the different page variations to a random sample of your visitors and update you as the experiment progresses.

Which Page Elements can be Tested?

What you test depends on the type of business you have and your individual situation. With Google Content Experiments, you can test things like page headlines, different images and icons, different text/content elements, calls-to-action, and even different page layouts! Google Content Experiments allows quite a bit of flexibility; however, it is limited to either URL destination goals or Event goals – i.e. as of now, you cannot use Analytics goals like time on site or pages per visit.

Testing Tips & Best Practices

As you test different web pages and content, it’s a good idea to follow some simple guidelines:

  • Don’t test too many elements at once. Doing so makes it difficult to determine which element or combination produced the best result.
  • Test pages with a decent amount of traffic. If you operate a small business, you’ll probably want to focus on getting more traffic before you run your first experiment. The higher your website traffic, the faster you can gain knowledge from your experiments.
  • Incremental tests stink. Everyone online has ADHD. Whatever you’re testing, make it BIG. Testing miniscule details isn’t likely to provide you with conclusive results.
  • A-B-T. Always be testing!

Will Google Experiments Hurt My SEO?

Many webmasters refrain from A/B testing because they fear it will negatively impact their SEO. Google does not view Content Experiments as “cloaking” and running experiments will NOT hurt your site’s SEO – as long as you follow a couple basic rules:

  1. When performing a content experiment, always use the rel=”canonical” link attribute on your test variation pages. This attribute sends a signal to search engines that they should index the content on your original page NOT your test page.
  2. IF you choose to leave your content experiment variation pages “live” after your experiment has ended (to make it so that people that have bookmarked your page don’t get a 404 error if they return post experiment), you’ll want to use a 301 (permanent) redirect to send visitors to this URL back to your original page. Alternatively, if you plan to use the test URLs again in the near future, you could use a 302 (temporary) redirect.

Learn more about Google Content Experiments and email us with your input after you run your first experiment!

Tags: , ,

Posted in Analytics, Analytics News, Analytics Product Updates, Analytics Tools, Google Analytics


Leave a Reply

Your email address will not be published. Required fields are marked *


Powered by WordPress. Designed by Försäkra Online.