I often see accounts set up in Google Analytics with only a single profile. This is really a shame. There is a lot of great information that is obscured with just a single, overall profile.

There are a several reports and metrics that you can’t segment within the Google Analytics reporting interface, such as

1. Funnel Visualization (Goals > Funnel Visualization)

2. Visitor Loyalty Reports (Visitors > Visitor Loyalty)

3. Ecommerce Reports

4. Number of Visitors

These are much more interesting (and useful) when you can compare them against different segments of your traffic or see trends within just one segment.

Funnel Visualization, for example, is a great report. You can see where visitors drop out of your funnel and where they go when they leave. But you can’t compare funnel behavior for New Visitors and Returning Visitors.

If you could look at them separately you would be able to tell that the change to your website that caused an increase in funnel exits is only affecting Returning Visitors. Maybe you moved a vital element to a different area of the page. There isn’t anything wrong with the new placement (New Visitors are doing just fine), except that your Returning Visitors are expecting it to be somewhere else.

But you don’t know any of that because you can’t look at the Funnel Visualization report by Visitor Type.

Or Can You?

By creating some extra profiles you will be able to isolate which group of visitors is responsible for the changes you see in your data.

Here are a few basic profiles you can use to look at segments of your website traffic separately. (Each of these links to an image showing the filter that you need to add to that profile.)

Only Direct Traffic

Only Organic Traffic

Only Paid Search Traffic

Only New Visitors

Only Returning Visitors

Creating Profiles

1. Click Add Website Profile

2. Select Existing Domain, Name Your Profile, Finish

3. Select Edit for the new profile, then Add Filter

4. Create the filter shown in the Links above for the profile you just created, click Finish

With these profiles you can compare your funnels for New vs Returning Visitors or your ecommerce reports for Organic vs Paid Search. Use your Organic profile to see if your SEO efforts are paying off, or your Paid Search profile to refine your pay-per-click campaigns.

Every time you are looking at Google Analytics and wish you could segment that report just one more level, you may have found another opportunity to create some extra profiles.

-John

We see it all the time — companies who just can’t figure out why their site doesn’t convert like it should.

We know first-hand that high conversion rate sites aren’t necessarily the prettiest sites. They often don’t have fancy graphics, knock-your-socks-off design or expensive flash elements. But they know what their users want and they give it to them — for example, they help users easily navigate to the right product; they give users the information they care about; they make it easy for users to do business with them.

Since it’s a Friday, I’m not going to tell you why you should do user testing on your site, or rant about common usability problems I encounter in performing user testing for clients. Instead, I’ll save the educational post for the work week and direct you to this funny (because it’s true) post, Don’t Hire An Ad Agency to Build Your Web Site. It says it better than I ever could!

The user defined segment in Google Analytics is a very underused feature. But those that do use it, are often
frustrated because you can only use it once on a site. However, there are situations where it can be safely and easily
used for multiple purposes. Here is a situation I encounted that makes an excellent example. It uses some
JavaScript to determine if the User Defined Segment Variable (_utmv) is already being used and decides what to do accordingly.

Quick Refresher:

The user defined segment based on the value of the _utmv cookie.
You set the value of this cookie with the __utmSetVar() function in urchin.js or with pageTracker._setVar() if using ga.js.
To view the user defined segment in your analytics, go to >Visitors->User Defined.

Background:

You currently use the _utmv cookie to segment your employees’ actions from other visitors, as well as exclude them from certain profiles.
For our example, we set the emplyee’s cookie value to “notrack”.
(Justin Cutroni has a great post explaining what this is and how to do it.)

Now, you want to track your members separately from other visitors. You’ve configured your login page to set the _utmv
cookie to “member” upon successful authentication. Now, they’ll be tracked as a part of the “member” segment for any pages they view.

The Problem:

The problem is that once your employees log in, their “notrack” cookie is overwritten with the “member” cookie.
On their next visit, it’ll look like they’re a regular member. (We don’t want that.)

The Solution:

With a tiny bit of JavaScript, we can easily keep the “member” cookie from overwriting
the “notrack” cookie when an employee logs in.

To set the _utmv cookie, we would normally just place the code below on the landing page after the user logs in.

- urchin.js

<body onload="javascript:__utmSetVar('member');">

- ga.js

<body onload="javascript:pageTracker._setVar('member');">

However, this would overwrite any _utmv cookie that was previously set. (ie, our “notrack” cookie)

We just need to add a short JavaScript function to see if the cookie already exists, and update our
onload event to call the new function instead of __utmSetVar.

    <script type="text/javascript">
    //<![CDATA[
    function tagMembers()
    {
    var allcookies = document.cookie;
    var mycookie = allcookies.indexOf("__utmv");
        if (mycookie == -1)
            {
                    __utmSetVar('member');
            }
    }
    //]]>
    </script>

* For ga.js, just replace “__utmSetVar(’member’);” with “pageTracker._setVar(’member’);”.

Your body onload event should now look like the example below:

<body onload="tagMembers();">

What does it do?

Instead of automatically setting the _utmv cookie every time, we check
to see if the cookie is already set. (It doesn’t matter what it’s set to. If it’s “member” already, we don’t need to set it again, and
if it’s “notrack” we don’t want to set it again.

Now, when your employees log into your site, they’ll continue to be tracked or excluded as such.

I hope this helps to expand everyone’s use of the _utmv cookie. With a little logic, you don’t need to be bound
to using the _utmv cookie for only one type of user. I welcome your questions, comments, and suggestions.

Coming soon…Adding logic based on the actual value of the _utmv cookie.