13 04, 2016

Four Ways that a Google Analytics Session Could Break

By | 2016-10-31T10:36:49+00:00 April 13th, 2016|Categories: Google Analytics|Tags: , , , |

Google Analytics provides great tools for data collection and reporting. One of the critical aspects of analytics for your business, though, is ensuring that you have clean data.

An issue that I come across occasionally is a lack of session continuity. This means that a visitor’s session on your website is split into two or more website sessions. When this happens, your website session totals, site usage metrics, and traffic source attribution are all filled with incorrect data.

Beyond those reporting metrics, eCommerce websites rely on data analysis that includes Multi-Channel attribution models, total sessions per transaction, and assisted conversions. Broken sessions makes this reporting unusable. This can hurt eCommerce website marketing, because so much is dependent on proper traffic source attribution. Picture this: A new visitor comes to an eCommerce website and begins shopping. At some point, the website session being captured by Google Analytics breaks, and a new session begins during the visitor’s same shopping visit. Once a purchase is made, the marketing channel that brought the visitor will be considered an assisting source, not the actual source. Depending on the website’s attribution model, this could hurt analysis that will lead to marketing budget decisions.

If you are not confident that your visitors’ website sessions are maintained from beginning to end, then you have some work to do.

Here four ways that a Google Analytics session could break:

1) Session Timeout

There is a 30 minute session timeout, by default. Once that mark is hit, the session ends. If a user kept a tab open in their browser with your website, and resumes activity in the browser, then a new session will begin. Google Analytics will note that it is the same user with a new session. You have the option of adjusting the timeout marker, anywhere from one minute to four hours. This can be done in the Session Settings section of Tracking Info configurations for your Google Analytics property (see screenshot below).


2) Missing Google Analytics Tracking Code

Every now and then, I come across a website that has not applied consistent Google Analytics tracking code across the entire site. Google Analytics data collection is based on its JavaScript being installed and readable. If a page on your website is loaded, and the JavaScript is not present, then there will be no data capture. Consequently, the previous pageview will mark the end of a user’s session. If the user clicks from the non-tracked page to a page with the Google Analytics JavaScript, then a new session will begin. Just like with the session timeout, your data will show one user with two sessions.


The best way to avoid missing code is to apply the Google Analytics JavaScript in your website template. Every time a new page is created with your template, the Google Analytics code will automatically be in place. If you use multiple templates on your website, be sure to apply the Google Analytics JavaScript to each template.

3) Cross-Domain Tracking Error

Anyone who intends to set up cross-domain tracking knows that the most valuable aspect of the setup is maintaining session continuity between the two domains.

The industry I most frequently see cross-domain tracking is Higher Education. Many institutions utilize multiple domains, and there is an obvious relationship between the content across those domains. Perhaps from the institution’s standpoint, content, data collection, and reporting are segregated by departments. However, the website visitors have a unified experience across all of the content during their website sessions.

Cross-domain tracking with Google Analytics can be tricky, which is why even Google recommends using the linker plugin to make life easier. Validate your tracking code setup in a test environment. If at all possible, never use your live environment to test your tracking code. If, while testing, you notice that the source/medium changes when clicking to the second domain, then there is an issue with the tracking code.

4) Conflicts with Other Scripts

A very talented JavaScript developer once explained to me that, with the broader use of scripts on websites these days, there is potential for some scripts to conflict with each other and cause one or both to break. The best way to catch this is to validate your tracking code setup in a test environment. Not only should you make sure that the Google Analytics reports are showing data, but you should also verify the appropriate Google Analytics cookies, and use your browser’s developer tools to check for any script errors.

Get More Reliable Data

Guarding against the above issues helps ensure that your website’s Google Analytics tracking is as accurate as possible. More specifically, these issues skew website session totals and sometimes source/medium attribution, which then skew conversion rates and other helpful KPIs. The goal of analytics is to make informed decisions that improve your business. Make sure your data is reliable, for that purpose.

13 08, 2015

Avoiding Google Analytics Tracking Issues

By | 2017-08-11T16:17:04+00:00 August 13th, 2015|Categories: Google Analytics|Tags: , |

I love that Google Analytics provides out-of-the-box tracking code when setting up a new property. It helps marketers plug & play for a quick turnaround. Unfortunately, this convenient feature can give you the false sense of reliable data. Just because data is populating your Google Analytics reports does not mean that the data is accurate and reliable.

Here are some Google Analytics Tracking Issues I have encountered over the years. Included with each one is a tip on how to avoid the issue.

Cross Domain Tracking Isn’t Working

When cross domain tracking is not properly configured, your report data will show two sessions instead of one each time a user clicks from one domain to the other. This inflates your website sessions, average pages per session, etc. Proper setup steps to avoid this include adding an additional line of code in the tracking script, as well as tagging outbound links from one domain to the other. Google provides some helpful information on cross domain tracking.

Why Am I Seeing Self-Referrals?

If you have subdomains that are included in your tracking, then you might be experiencing this. A quick look in the Referrals report sometimes shows visitors being brought to the domain from itself. To avoid self-referrals, be sure to add your domain (without the www) to the Referral Exclusion List.

No Analytics Data for Some of My Pages

Perhaps you have analyzed the content reports and seem to think that some pages are not being hit. This could be happening for many reasons. For now, we will focus on tracking code implementation. I have worked with some websites that utilize multiple templates. In my opinion, it is best to apply your Google Analytics tracking code to the header in your content management system (assuming that each template uses the same header). This will ensure that adding a new template will not result in missing tracking code.

Referrals From My Test Environment

This is more of a web development issue than a Google Analytics issue. Any time this is seen, my first instinct is to check all links in the test environment. They should be relative, meaning that they will link to the test version of the destination URL. When the destination URL is an absolute references, you could see website sessions sent from your test environment.

Analytics Pageviews of My Test Pages

This issue is not easily caught. For one, it is easier to catch when you add a filter to display the full URL of each page on your site. There are two options to avoid this. First option is to add a filter that excludes the hostname of your test environment. The second option is to create a second property in Google Analytics for your test environment and have the UA-ID automatically switch between the two environments.

Exclude Internal Traffic

Similar to the previous issue, simply viewing Google Analytics reports will not necessarily tell you if your office’s activity on the website is skewing the report data. Upon setting up your Analytics property, it is imperative to have at least one reporting View with an IP address exclusion filter. If necessary, check with your office’s IT department to determine if you need to exclude one IP address or a range of IPs.

Google Analytics Goals Not Working

So, you are receiving contact forms in your inbox, but Google Analytics is not showing any goal completions. The most common way this issue occurs is by not utilizing the proper setting with the destination URL. You have the option of selecting Equals To, Begins With, or Regular Expression. If your confirmation page ends with a variable query parameter, then you cannot use Equals To. If you have configured your content reports to show the full URL, then you cannot use Begins With and start the destination URL with a forward slash. If you elect the Regular Expression option, familiarize yourself with regular expressions, and test the regex first.

Test Your Setup

Above all else, I cannot stress how important it is to test your Google Analytics tracking code setup before adding it all to the live version of the website. Many marketers make the mistake of using the live website as the testing ground. This is a major cause of Google Analytics tracking issues. Set up everything in your test environment first. If you are implementing event tracking, custom dimensions, or any other advanced feature, the testing time will prove valuable. Seek out Google Analytics support when you know the tracking and data needs are beyond your skill set. Give yourself enough time to set up everything in the test environment. It is better to work through these issues there than after the boss asks for a report.