IWMW 2016: Google analytics masterclass
The following post is based on a plenary talk and masterclass given at the 2016 Institutional Web Management Workshop (IWMW) by Martin Hawksey. A video of the plenary talk is available and is well worth watching. There are also slides available from Martin’s site. The following notes are my highlights from these presentations.
Pageviews
Pageviews are no longer a meaningful measure of traffic to a website as pages are longer and web applications like Slack have content that is not tracked in Google Analytics. It is better to understand user engagement via event tracking to see how a user moves through a website.
Google doesn’t allow personal information to be passed to Google Analytics, but it is possible to send an encrypted client ID to identify a user across all devices.
Sessions
A session is the activity from a specific user or browser within a 30 minute period. However, the session will end at midnight, so a user comes to a web page at 2355, the session will only be 5 minutes long. Different timezones will also affect this. If you return via another channel the session starts again, or if a new tab is opened, it is not tracked.
Google Analytics relies on cookies and JavaScript, so ‘drop-in’ sessions could be due to the code which they rely on not working.
This means that you should stay away from aggregate metrics like bounce rate and pages per session. Instead, we need to look at the funnels i.e. the pathways to a web page, that produce these outcomes.
Google tag manager
Google tag manager allows you to insert code into a web page via the Google tag manager interface. This allows endless possibilities for injecting code into a page to track user behaviour as well as potentially modifying the behaviour of the page itself. For example, it can detect if a user is on a mobile device and use this to track specific events. There are online courses available on tag manager fundamentals. The following are examples of how Google tag manager can be used.
Bounce rate vs. SERP bounce time
Bounce rate is when a user comes to a web page and then leaves to go to another page outside the website during a session. An alternative approach is to determine how long a user stayed on a landing page after arriving from an organic Google search and then going back to the Google search engine results page (SERP). The recipe to do this using Google tag manager is described in an article by Simo Ahava.
Track content engagement
Google tag manager can be used to determine how much time users are actually spending engaged with content on a page. Simo Ahava approaches this by treating content as a product. The following video from Ahava gives an overview of this concept. Ahava has written a recipe for doing this with Google tag manager.
Simo Ahava: Meaningful Data from Reaktor on Vimeo.
Automatic reporting
The Google Analytics add-on for Google Sheets allows data to be combined from different data sources for reporting via email. Martin has developed a method for combing data from Twitter and Google Analytics to run automatic collection of search results from Twitter. This would allow you to find top referrers for tweets.
Another example is the UK Cabinet Office created a dashboard to display data from Google Analytics in a user-friendly format on a web page.
Conclusion
Design by data is one of our key principles. The next step is to explore how we can use Google tag manager and automatic reporting at the University of St Andrews to provide meaningful data that can help us improve digital services.