Step by Step Guide to Google Analytics Content Experiments

A recently introduced new feature of Google Analytics is called Content Experiments. With this new feature you can easily test multiple variations of a page to see which version works best on your website or webshop to improve the number of conversions.

Simply said: this way you serve version A to the first visitor and version B to the second visitor. With a goal set you can then find out if for example you have to place that button to request information on the left or on the right, or maybe it’s not a good idea to use big images at the top of the page.

The new feature is for a great part based upon Google Website Optimizer which was always a stand-alone product, for this reason Website Optimizer will be shut down August first. We’ve discussed this at our hangout last week, but let’s take a look into this ‘new’ feature.

Start experimenting

To start experimenting within Analytics go to the website you want to test with and click on the left on “Content” and then “Experiments”. On the page that opens fill in the URL of the page you want to test and hit the big blue button. Setting up an experiment consists of four different steps.

Name and URL’s

After clicking the button on the next page you’ll first give the experiment a name. Below the name you have to fill in the URL and name of the original page, after doing so Analytics will generate a screenshot on the right. Below the original you can fill in the URL of the first variation of a page, so in reality this page is a completely new page on a different URL then the page were the experiment is tested on, when the experiment is live Analytics will combine these.

Analytics currently supports experimenting with five different variations. Hit next to continue.


After filling in the URL’s it’s time to set a goal. At this moment only URL destination and Event goals are supported. If there are no goals defined in the profile then click the link ‘Setup goal now’ to add a new goal. After that you can decide how many of new visitors must be included in the experiment, normally this is 100% but you can choose from 1%, 5%, 10% to 25% to 50% to 75%.

Last click the option ‘Rewrite variation URLs to original in Content reports’ to consolidate all traffic to your original and variation pages under the URL for the original page in your Content reports.

Add the code

In the third step it’s time to add the code to your website. Here you have the option to do this yourself, which is of course very cool to do, but there is also an option to send instructions to your own webmaster. When selecting this last option you will get a standard email which can be changed, fill in an address and the webmaster who receives the message can click on a link to open a page with the instructions and the code to place.

If you decide to add the code to the page yourself then click the first option. The code has to be placed after the opening head tag at the top of the original page. Of course, also make sure all pages (original and variations) have the normal Google Analytics code.


Then click Next to validate the pages, Analytics will then check your pages for working code. If the test is succesfull hit the Excellent button to continue. In the last step you will need to confirm all settings and then you can start experimenting for real or save the experiment for running it somewhere in the future.

When the experiment is started it will take 24 hours up to 48 hours before the first data can be seen in Analytics. If there’s enough data Analytics will declare a winner and show which page performs best. Note that all experiments are limited to three months. Later on the experiment can be opened by going to Content and then Experiment again. After opening the experiment you will see the conversion rate, numeber of visits, hours of data and the chances to beat original page.

It’s a very easy way to perform experiments on a website, so happy experimenting!

About Michel Wester

Michel Wester is the owner of the Dutch website, started in 2004. WebSonic mainly focuses on news about Google, Twitter, and Facebook. Next to the news the site has a lot of tips for PC users and starting web developers.

23 thoughts on “Step by Step Guide to Google Analytics Content Experiments

  1. Thanks for a nice article Michel,

    I think the fact Google have decided to put this functionality as part of Analytics is great news. The massive shame is that it only has the ability to perform split tests (A/B) and not the multi-variate capabilities we have become accustom to.

    Hopefully this will be fixed soon

    Russell McAthy
    Digital Marketing Consultant

  2. Hi Michel and Russell, we have been hearing this … the missing multivariate testing features and that is why we launched 4 days ago. Seamless Google Analytics intergration, revenue tracking, multivariate testing, url split testing, visual A/B testing and sure easy (one tag site-wide)… and multiple goals 🙂

    Have fun try it out free.

    PS: yep I am pitching I am the proud co-founder 🙂

    1. just 1 question has anyone seen any data getting captured in experiments?? I have been using it for weeks however not a single visit reported by content experiments report… main site is seeing visits including targeted pages. However no data in content experiments. 

      I replicated the setup for public usage/testing at original page – 1 – 2 – –

      1. Pallav, sure there is no filter active on the account? That can stop data…
        It also takes 2 weeks before first data will show in Google Analytics CE.

        Make sure you have turned on notifications, they might send a note when there are problems. If nothing works… sign-up for on 30 day trial 🙂


  3. The major problem that I understand from here is that it will only work on Google trafic, and not direct trafic, whereas the google website optimizer worked for it.

    Did I understood well ?

    1. Hi BH SEO,

      Nope split testing will work for all traffic. So no problem getting started. I do want to invite you to if you ready to grow beyond split testing and start MVT and A/B testing (visually). We also integrate well with Google Analytics are launched last week.

  4. Something messed up the urls
    I replicated the setup for public usage/testing at original

  5. Hello:
    I would love to do this experiment, but the blue button isn’t highlighted so it’s not clickable.  There’s a message underneath that says “Only administrators can add new experiments”.  I don’t get it, I am the administrator…..


    Thanks for sharing this great information…..

  6. Hi

    Very nice article. It helped me a lot understanding content experiments.

    But I have a question.

    My goal type is currently URL Destination.

    And we push conversions manually to analytics after phone call matching usually it takes one day.

    We use this code for pushing conversions.

    var _gaq = _gaq || [];

    _gaq.push([‘_setAccount’, ‘XX-XXXXXXXX-X’]);

    _gaq.push([‘_setAllowAnchor’, true]);

    _gaq.push([‘_trackPageview’,’/ ‘]);

    What I want to know is how do I push a conversion manually for a specific variation.

    Thanks in advance for help.

  7. Hi,

    Great article, totally new to this so this is a bit of an elementary question. Once you have gone through the steps above what is the process involved that routes traffic between the control and experiment pages. I.e. what/where/who does the following (assuming that there are 5 variations AND the process operates on a round robin basis – is the second assumption correct?):

    Visitor 1 -> control page
    Visitor 2 -> experiment page 1
    Visitor 3 -> experiment page 2
    Visitor 4 -> experiment page 3
    Visitor 5 -> experiment page 4
    Visitor 6 -> experiment page 5
    Visitor 7 -> control page (is this how it happens – round robin style back to the control?)
    and repeat…



  8. I’m wondering how to test multiple variations of my homepage. Most of these bloggers don’t write in English, or they’re very lazy. Can you explain how to test multiple variations of my homepage instead of just the original and one variation? That seems to be the only choice that Google Content Experiments offers, but some of these Geak-ese speaking bloggers suggest otherwise without explaining in English how to do this.

    1. If you respond, I’m not sure whether I’ll get it. There doesn’t seem to be a follow link for comments. I really want to know. Google makes it really difficult. These bloggers don’t help either. They only speak to fellow nerds.

Comments are closed.