We use Google Analytics on most of our sites. It’s free (within limits), it’s easy to implement and there aren’t many things it can’t track.
The login details for the shared Google Analytics account are kept in 1Password, search for “Google Analytics”.
By default, pageviews are the only user interaction that Google Analytics will record.
If there are specific events you want to track that Google Analytics will not track by default (e.g. clicks on links to external sites, file downloads), you will need to implement a custom event, as per Google’s instructions.
It’s important to note the section on handling timeouts. If you don’t do that, the links you are tracking won’t work if GA doesn’t load (e.g. if the user has a browser extension that blocks trackers).
There are a few basic steps it’s worth following for any site you’ve set up in Google Analytics:
- Set up an unfiltered view that records all traffic, including spam. Any new view you create is unfiltered by default. You won’t use this view for most analysis, put it provides a control you can compare against, and a backup in case you get the filtering wrong on other views.
- Set up a “main” view that has Google’s “Bot filtering” activated (found under Admin > View Settings). This should automatically remove the majority of spam traffic.
- You may also want to filter out internal traffic on the main view (e.g. from within the dxw office, or staging/localhost versions of the site). You can do this by IP (e.g. using dxw’s external IP to filter out office traffic) or hostname (e.g. the staging site URL to filter out staging traffic).
- To avoid double-counting page views that differ only by query string, follow Google’s advice to build a list of query parameters your site receives, then blacklist the ones you don’t care about using the field in View Settings.
- Set up Google Search Console (formerly Google Webmaster Tools), and give Analytics access under Acquisition > Search Console. This will allow you to see the search queries people are using that return the site in Google results.
Take some time to think about what activity on the site you might want to track. If you do not record a user action at the time it happens, it is lost forever, so the more you can set up in advance (e.g. goals for transactions such as newsletter sign-ups, recording of on-site search terms), the better.
Understanding Google Analytics
GA uses a lot of jargon. It also has a tendency to present its data as absolute fact, even when there are all kinds of assumptions or complexities that underlie it. Some good resources for understanding it:
- Average Time on Page: Understanding Google Analytics’ Average Time on Page
- Bounce rate: Adjusted Bounce Rate (does a good job of explaining the limitations of bounce rate, and offers a workaround for them)
- ‘Direct’ Traffic: Guide to Direct Traffic
- “Next Page Path”: Misunderstood Metrics: Next Page Path