#custom variable

[digitalmeasure]

Prevent Employees from Appearing in Analytics (update for Universal Analytics)

As promised, here's the article with a walkthrough of how to exclude employee visits, such as content authors and business users, from your Google Analytics Universal Analytics profile. This article is a follow up to an earlier one written on excluding employee visits for the classic (async) version of Google Analytics. Both articles address the need through a persistent Web browser cookie and a GA filtering rule.

Note: This article is written for Google Analytics Universal Analytics. The code and configurations will not work with async. See the earlier article on excluding employee visits from Google Analytics async.

Google frequently changes the interfaces for both administration and reporting. This walkthrough is accurate as of the writing of this article. Regardless of where something is found in the interface, the terms and process mentioned here should remain the same.

We'll follow the same three step structure from the async article. I'll glaze over any elements covered in the previous article so we can move directly to the implementation. That information is available in the async article.

Step 1: Assign and Set Custom Dimension on your site’s domain

Step 2: Test before implementing in production profile

Step 3: Configure filter within production Google Analytics profile(s)

Step 1: Assign and Set Custom Dimension on your site’s domain

Assign Custom Dimension (Server Side)

Before we set the variable value, we need to set the variable name. GA's async custom variables names are set client side (the variable we set in the previous post is actually an even older variable from what I'd now call GA antique synchronous tracking, and doesn't even offer a variable name). In GA's Universal Analytics, custom dimension names are now set server side. This makes management of these names much more streamlined since variable names are stored in one central place and you can avoid digging into the code to change them. Aside from this difference, GA Universal Analytics offers 20 custom dimensions, up from 5 in async (and one in antique sync). The rest is pretty similar, with hit level, session level and user level variables.

click on "Admin" in the top right.

under the "Property" column, expand "Custom Definitions"

click on Custom Dimensions

click on New Custom Dimension

For Name, enter "User Type",

For Scope, select "User"

Check Active

Save

Google Analytics Custom Dimension: in the Web Property that all the Profiles belong to, set a new Custom Dimension for User Type.

After you've saved the new Custom Dimension, note the index number assigned to this custom dimension. If you haven't set one before, this should be 1.

Google Analytics Custom Dimensions Table: after saving the new Custom Dimension, you will see it appear in a table with the Custom Dimension Name you assigned and the dimension's Index. The Index number is the slot you'll use in when setting the Custom Dimension Value on the client side.

Set Custom Dimension Value (Client Side)

GA's async custom variables values are set client side. This is no different in GA's Universal Analytics, with custom dimension values set client side on the page.

Create an orphan Web page hosted on your site's domain with your site's standard analytics code. For instance, http://foo.example.com/exclude-visitor-analytics.html. If you'd like, add a meta robots noindex tag to the page's head so it doesn't accidentally get crawled.

Ideally the dimension is set before ga('send', 'pageview'); rather than after it. If it's sent afterwards, the dimension is submitted with the next pageview rather than this current pageview.

ga('create', 'UA-XXXXXXX-X'); //set dimension prior to sending the pageview like so ga('set', 'dimension1', 'employee'); ga('send', 'pageview');

For simplicity, you can choose to add the code after sending the pageview by adding this to the HTML code of the orphan page, where dimension1 is the custom dimension slot you'll be using:

<script type="text/javascript"> ga('set', 'dimension1', 'employee'); </script>

If you already have custom dimensions, set an empty slot and document it where appropriate. Keep this slot exclusive for employee tracking.

Step 2: Test before implementing in production profile

Test with a separate profile to track only those pageviews containing the new custom dimension.

In your Google Analytics administrative interface, access your profile:

click on "Admin" in the top right

under "View (Profile)", expand "Filters"

click on New Filter

for Filter Name, enter "Exclude Employees"

select Custom filter

for Filter Type, select "Exclude"

For Filter Field, find "User Type" under Custom Dimensions

For Filter Pattern, type in employee

Set Case Sensitive to "No"

Save

Google Analytics Profile Filter: configure the production profile to exclude pageviews from any visitor that has the User Type custom dimension (from Step 1) set to "employee".

Step 3: Configure filter within production Google Analytics profile(s)

Now that you’ve tested the custom variable, you’re ready to move it to production. In your production Google Analytics profile(s), add a new filter to remove visitors that carry this custom variable. Follow the screenshot in Step 2, but select “Exclude” instead of “Include”.

Google Analytics Profile Filter: configure the production profile to exclude pageviews from any visitor that has custom variable set to “analytics-exclude”.

[digitalmeasure]

Clean your Analytics by Excluding Visits from Employees

If you wish to keep your site's analytics clean, you should filter out employees that visit your site. This allows you to better focus on your site's legitimate visitors, the customer.

A simple way to exclude content managers and business users from your site's analytics account is to have visitors self identify themselves for exclusion.

In Google Analytics, the internal visitors will view a hidden page on your site. Once they access the page, they're assigned a custom variable in Google Analytics for this and all future visits to the site. Google identifies the custom variable through a cookie stored in the visitor's Web browser.

You can do whatever you wish with visitor data containing the custom variable, including removing those visitors from your site's analytics. You can even create a separate profile to track only their visits. This article is wriitten with async code in mind. A separate post is available for instructions on preventing employees from showing in your Google Universal Analytics reports.

Before we get you on your way to cleaner, better data, you'll need to know which version of Google Analytics your site is running. Google Analytics is undergoing an overhaul as it upgrades from "classic" analytics, known as async, to the shinier, more flexible Universal Analytics. The two flavours of Google Analytics use the same reporting interface. The two versions are incompatible, so if you're running async, you need to use async code, and likewise for Universal Analytics. Google has promised a forthcoming upgrade path for users of async so they can move to Universal Analytics. At the moment, the two are completely incompatible. If you're using async and satisfied with it, you should stay with it until Google releases an upgrade path.

Here's how to tell which version you're running. There are a plethora of other ways, this is probably the most foolproof way.

Sign in to your Google Analytics account and access the Web property you're working on.

From the top right, click on "Admin".

In the column for your Web Property, click on "Tracking Info", then from the dropdown, select "Tracking Code"

This section will tell you which version of Google Analytics you're running.

classic async - look for text that asks you to upgrade, or that tells you you're running async code for this property.

Universal Analytics - look for text that tells you you're running Universal Analytics tracking code for this property.

Note (added November 16, 2013): The remainder of this article is written for Google Analytics classic code (async). The code and configurations will not work with Universal Analytics. Here's an article to walk you through excluding employee visits from appearing in Google Analytics UA.

Read the notes below for further details about exclusding employees from your Google Analytics classic async, and see the comparison below between the custom variable method and another method of excluding employee visits.

Step 1: Create Web page on your site's domain

Create an orphan Web page hosted on your site. The orphan page must be located within the domain of your site and it must contain the standard analytics code used within your site. For example, if your site is foo.example.com, the orphan page you create needs to be in foo.example.com, such as http://foo.example.com/exclude-visitor-analytics.html.

Ideally the dimension is set before _gaq.push(['_trackPageview']); rather than after it. If it’s sent afterwards, the dimension is submitted with the next pageview rather than this current pageview.

_gaq.push(['_setAccount', 'UA-XXXXX-X']); //set variable prior to sending the pageview like so _gaq.push(['_setVar', 'analytics-exclude']); _gaq.push(['_trackPageview']);

For simplicity, you can choose to add the code after sending the pageview by adding this to the HTML code of the orphan page:

<script type="text/javascript"> _gaq.push(['_setVar', 'analytics-exclude']); </script>

As a precaution, you might want to ensure that this orphan page is not crawled by search engines should someone accidentally link to it from a publicly accessible Web page. Add this page to the exclusions list in your robots.txt file or add a robots meta tag to the head of the Web page with noindex nofollow.

Step 2: Test before implementing in production profile

It's always a good idea to test that changes are implemented as planned prior to deploying them in the production Google Analytics profile. The test profile is also useful if your site receives a lot of visitors, as your analytics reports will be noisy.

You can test by creating a new profile for your Web property. This new profile will track only those pageviews containing the custom variable. Only visitor that have viewed the orphan page will have this custom variable.

Note: avoid creating a new Web property when you create the new test profile. Your test profile must use the same UA tracking number used in your production site.

Create a new filter for the test profile following the screenshot below.

Google Analytics Profile Filter: configure the test profile to include pageviews from any visitor that has custom variable set to "analytics-exclude".

You will not need to apply other filters to this profile.

Real time analytics will tell you within seconds if everything is working correctly. In the test profile, go to "Reporting", then "Real-Time", then load the orphan page from your site. Visit any other page on your site that contains your standard analytics tracking code. You should see that your pageviews show up.

Step 3: Configure filter within production Google Analytics profile(s)

Now that you've tested the custom variable, you're ready to move it to production. In your production Google Analytics profile(s), add a new filter to remove visitors that carry this custom variable. Follow the screenshot in Step 2, but select "Exclude" instead of "Include".

Google Analytics Profile Filter: configure the production profile to exclude pageviews from any visitor that has custom variable set to "analytics-exclude".

Notes about the Custom Variable

The custom variable you set is stored as a cookie on your browser named __utmv.

Visit this page on every browser you use, including desktop browsers, tablet browsers and smartphone browsers.

The __utmv cookie expires after 2 years.

If you revisit the page any time before the cookie expires, it will renew the cookie for another 2 years.

The __utmv cookie is set to the same domain that the rest of your analytics is set to. This only matters if you've specified _setDomainName in your analytics script. If you did specify a domain in  _setDomainName, then that is the domain scope for your cookie.

__utmv is a custom variable that is officially deprecated. It is still understood by Google Analytics. The custom variable is still widely used, the most popular use is likely the use described on this post, excluding visitors.

If you clear your cookies, you'll need to revisit the page.

You can only set one custom variable through this method. Set up to five custom variables through the new method allowed by Google Analytics, _setCustomVar.

Another Method to Remove Employee Visits

You can instead filter traffic by IP address. Both the custom variable method and the IP method have advantages and drawbacks.

The custom variable method requires users to visit the orphan page on their Web browser. It's more difficult to communicate this requirement and remind staff to visit the page if the content management team is widely distributed, such as within a university setting.

The IP address method requires that the business have a range of static IPs. Those IPs are filtered out within analytics reports. It will not work if corporate machines use dynamic IPs. If employees work remotely through a VPN, those employees should be using the same static IPs as they would at the office. However, any device accessed outside of work that's not on a VPN will not be excluded through IP filtering. This method also is not suitable for most home businesses, sole proprietors, site owners and bloggers that use consumer ISP plans, as those plans usually provide dynamic IPs. Static IPs cost a premium so most home consumers and many small business customers of ISPs tend to stick with dynamic IPs.

Further thoughts for Developers

It's worth noting that you must push the variable to your analytics through gaq. Google creates a cookie __utmv. The value of the cookie is set to your domain hash + .analytics-exclude For my site, it's 241544733.analytics-exclude

I tried setting the custom variable without using gaq push out of a curiosity to see how the variable is set. When you use regular JavaScript to set a cookie, it's not reported to Google and Google Analytics doesn't know the cookie exists. If you're curious, you can attempt it yourself by using jQuery with the jQuery Cookie plugin. Setting your own __utmv cookie takes three lines of code (instead of .digitalmeasure.me, use your site's domain):

var utma_hash_value = $.cookie('__utma').match(/(^[\d]+)/i)[1]; var exclude_visitor = utma_hash_value+'.exclude-analytics'; $.cookie ('__utmv', exclude_visitor, { expires: 730, path: '/', domain: '.digitalmeasure.me' });

Firefox has a great extension for developing and managing cookies called the Web Developer Toolbar. You can use it to inspect cookies, clear them by session, path or domain, and edit their values. Chrome has similar extensions.