Device/Browser Inventory for Responsive Testing (Custom Report Template Provided)

Custom reportview at full size

This custom report drills down from Device Category to Browser Version

Before rolling out a responsive or adaptive Website update to your users, it’s important to test the design changes on the same devices and browsers that your visitors are already using.

You can click the following link to import a basic device/browser custom report into one of your Google Analytics views:

https://www.google.com/analytics/web/template?uid=ev-n199TST2suk9H3E2Atg

If you are unable to source physical devices to test on, a fallback option is the device emulator built into Chrome or other available device emulators.

Note that in addition to the Users and Sessions metric, the custom report also includes Bounce Rate, just to demonstrate how current users are (and aren’t) interacting with your site across different devices and browsers.

In fact, if you have not yet incorporated any responsive/adaptive capabilities into your website, you could easily add metrics such as goal conversion rate and return visitors to the custom report to assess the need for a mobile strategy in terms of user counts and associated performance metrics. (Hint: the need is probably more urgent than you may think.)

Please share this post:
Facebook Twitter Plusone Linkedin Email

Spot-Check Your Google Analytics Events in Different Browsers

Browser and OS Reportview at full size

Using your Browser & OS report as a reference and Real-Time reports for testing, you can verify that your events execute in different browsers.

Different browsers can behave quite differently. When you add code for an event or a virtual page to your site, it can be helpful to verify event execution as follows:

1. Refer to the Browser & OS report report.

2. Select Browser as the primary dimension and Operating System as the secondary dimension (or vice versa) to display the most-used browser/OS combinations for your visitors.

3. Preferably in a testing/development environment for your website, and using a Google Analytics profile that is filtered to display activity only for the hostname of the testing/development environment, visit the site and perform the activity that is coded to generate the event.

4. Check the event in the Real-Time > Events report.

The following browser/OS combinations account for much of the traffic to many websites:

• Firefox on Windows

• Firefox on Mac

• Chrome on Windows

• Chrome on Mac

• Internet Explorer on Windows

• Safari on Mac

• Safari on iOS (iPad, iPhone, and iPod Touch)

As explained above, refer to your Browser & OS report to view the browser/OS breakdown for your own visitors.

Please share this post:
Facebook Twitter Plusone Linkedin Email

Isolating an Error to a Specific Browser Version

view at full size

Goal 1 Completion Rate for visitors using Firefox 16 was 92.69% lower than the site average.

Here’s a tip relating to a Javascript error that was occurring in the browser and therefore not detectable as an error page redirect originating on the server, which was discussed in a post earlier this week.

As admin for the Google Analytics Test project, I received detailed and very helpful email yesterday from a participant stating that he was encountering an error on the setup page of the site while using Firefox 16 as his browser.

Once I duplicated this error in my own installation of Firefox 16, I logged into GA, drilled down to Firefox 16 in the Browser & OS report, and saw that the Conversion Rate for Goal 1 (beginning the test) was only 2.90% for the previous week. Goal 1 Conversion Rates for earlier versions of Firefox, and for other browsers, were much closer to the site average of 39.64%.

I also saw the number Firefox 16 visitors likely impacted by this error in the previous week: 69.

Fortunately, the outstanding Firefox maintenance team at Mozilla was able to provide a quick fix for www.googleanalyticstest.com. (The problem: incompatibility of Firefox 16 with an old version of jQuery that I was still running.) And thanks to Google Analytics, and to the participant’s email, I had been able to isolate the error to a specific browser version and to determine how many visitors had been affected.

Please share this post:
Facebook Twitter Plusone Linkedin Email