Optimize Your Way to Riches: How to Use Google Analytics Content Experiments

How good is your website? Could it be better?

What are you doing right with it? What are you doing wrong?

What if you could make a simple change that increases the number of signups you get to your consulting newsletters by 25%?

Most people have no clue.

Their website, sign up page, sales letters and the like, once created sit there with little adjustments for months and often years.

What if your website could get your phone to ring more often?

What if you could make a simple change that increases the number of signups you get to your consulting newsletters by 25%?

And what if you could increase the conversion rate of your website by 10%, without spending a dollar more on traffic?

All three sound good, don’t they?

Well, you can achieve them. And doing so isn’t as difficult as you may think…

By using Google Analytics free Content Experiments service you can easily test two pages or several variations of a page against another and track the results.

Here’s how to use this great free service:

Until recently all A/B and Multivariate tests were done through Google’s Website Optimizer (GWO). This August Google has shut the service down and instead added very similar functionality into Google Analytics.

The positive is that it’s now under one roof, though setup can take a few minutes getting used to if you’re used to GWO.

1) Setting Your Goals

The first step in the process is to setup a Goal within Google Analytics.

To do that, click on the ‘Admin’ tab on the right side once you’ve logged into Analytics.

Next, under ‘Profiles’ select the ‘Goals’ tab.

Now you want to click on the ‘+Goal’ link.


(Analytics allows you to create 4 sets of 5 goals. I have several goals setup for the site I’m working on here, so you’re looking at the 4th set…if you’re new to this, you’ll see it as Goals (set 1).

To setup your goal you need to decide the following:

1) Give your goal a name.

2) Select ‘URL Destination’ (this means you will measure a completed goal when someone reaches the page/URL you’ve specified.

3) Enter your ‘Goal URL’ (for example, if you are measuring newsletter signups, you will put the URL of the page that the user lands on after they’ve submitted the form).

4) Match Type is next (this gets a bit geeky…but if the URL may change based on how people arrive, it is best to set it as a ‘Head Match’. That way if someone comes to /signup.html or /signup.html#fromadvertisingvariables both will count as goals.

5) The last part that is important is selecting ‘Use funnel’. This will get a bit more complicated (and it’s optional) but setting up a funnel will show you how people are navigating as they work towards your goal. For example, you can set this up to see how people move from your home page, to your signup page, and then to the thank you page. And see how many people move through your funnel and where people drop off.

Done that? Awesome, you’ve now setup your first Goal!

2) Create Your Variation

The next step is to make sure you’ve created the variation that you’re going to test against the original (your current version). You might test a different button color, headline, image, background, copy, etc…the list goes on.

So the page you’re testing might be www.yourdomain.com/newsletter.html and your new variation will be at page www.yourdomain.com/newsletter2.html

Be sure that you’ve tested your new page to make sure everything is working fine.

Once you have that new page ready…we can move forward.

3) Create Your Experiment

Now, to get back to your main Analytics screen click on the ‘Standard Reporting’ link in the top navigation.


The next step is to create your experiment.

On the left side menu you’ll see the ‘Content’ tab. Click that and then click on ‘Experiments’

Now click ‘Create experiment’ (if you’ve never created an experiment in your account before, you won’t see this button…and will be asked to enter the URL of the page you want to test as below…)


Okay, now you need to enter the URL of the page you want to test (this is your original page, eg. www.yourdomain.com/newsletter.html).


The first step in this process is to fill in the following:


Experiment name – pretty straightforward. I always find it helpful however to add the current date to the experiment name so I can easily see how long it’s been running.

Under ‘Original Page’ you’ll see the URL of your original page already entered in the form field. You can name your page or just leave it as ‘original’.

Next is the Variation.

As mentioned above, here you’ll enter the URL of the new page you’ve created. (…/newsletter2.html)

Again, you can give a name to this variation. I find it helpful to name the variation with whatever the main change I’ve made is. (If I’m testing a new headline, I’ll write something like “Emotion-based headline”)

Getting closer!

You’re now at step 2.

Now, select the Goal that you setup in Analytics which I explained earlier.

There are more advanced options here like splitting a percentage of the traffic to see the test. In general and for most smaller sites you can leave everything at the default.

Go to the next step…

At step 3 you have two options.

You can select to add the tracking code to your website yourself, or have your webmaster do it for you.

If you have any knowledge of html (even very basic) you’ll find this process pretty easy. Just follow the tips provided when you select that option.

Or you can enter the email address of the person you want to have add the code to your site for you.

Once the code has been added you can click ‘Next’ to validate your code.

The system will tell you whether the code has been installed correctly or not.

4) Verify and Start!

The last step, once you’ve verified that your code is installed correctly is to start the experiment.

Then sit back and watch the results come in.

(Results image above from lunarmetrics.com)

Tip: Don’t end an experiment too soon. Often you’ll see one of your pages doing much better in the early days, however, as you continue to let the experiment run the other page will actually outperform. This is called waiting to get statistically valid data. I may not be an economics or statistics expert….but I can tell you it’s important for you to have valid data before you make a decision.

If your Variation page outperforms your Control page, make your Variation page the new live page on your site (replace the Control with the Variation). Now your Variation will actually become your new control when you set up your next experiment and try to test a new variation to continue achieving better results.

Testing and tracking is what separates the real marketers from the wannabes.

Get testing.

Please Share This Article If You Enjoyed It:

  • This is helpful. Thank you.

    Here is my question: Does the Goal only get updated if the visitor doesn’t leave the website?

    I mean, if I am testing my newsletter signup page, and the person enters their email address, and then closes the browser. Then later, they go check their email, see the “confirm subscription” link and click on it, this then takes them to the page I set for my “goal.” But the goal was not completed during the initial browser visit, and may have been completed on a different computer even.

    • Jeremy – if they are on the same computer, typically yes. A ‘cookie’ is placed in their browser. So if they use a different browser or computer it won’t count. There have been many changes since I wrote this article. Best to look at the Analytics website or check out visual website optimizer or optimizely.

  • Samantha Avelon

    I got fed up with trying to understand Google Adwords PPC so I just called results driven 325-446-1507 and got a free audit of my campaign to help.