As of April 2014, Google announced that the latest iteration of its ubiquitous Google Analytics web analytics platform was ready to be brought out of beta and into the limelight. Unsurprisingly, it didn’t exactly set everyone’s imagination afire here in not-so-analytics-obsessed Malaysia.
For the typical business user, the first difference you would spot is the change of terminology – ‘Visits’ are now known as ‘Sessions’ and ‘Unique Visitors’ are now known as ‘Users’. As you will soon find out, this change in terminology is part of a move towards a more integrated understanding of user behavior.
We’ll go through some of the main differences between Google Analytics (GA) and Universal Analytics (UA) after the jump.
Different Tracking Codes
To determine which tracking code you or another website is using, you can follow the steps below:
- Go to your website
- Inspect element
- Check your tracking code, which can typically be found between the <head> tags. You might have to click and open the various <script> tags until you find it.
There are two different versions of ga.js – the traditional synchronous version and the asynchronous version.
It is now common to find the tracking code in the <head> section of your website after Google released the asynchronous version of ga.js. The asynchronous code is an improvement over the traditional ga.js GA code in that it allows faster browser execution and doesn’t block the page from rendering.
With the traditional GA code, you had to wait until it loaded fully before other elements such as your images or stylesheets could be loaded. This was why the traditional ga.js had to be placed in the <body> section of your website while the asynchronous version could safely be installed in the <head> section (I’m sure you already know this, but for those who don’t, your browser loads the code in the <head> section of a website first, then the <body>).
The much older version of Google Analytics is now popularly known as Urchin Analytics as it uses the urchin.js library. If you’re using this please upgrade soon – kittens die every day you’re using urchin.js.
How GA & UA Collect Data: Cookies & the Measurement Protocol
For the uninitiated, cookies are text files that are usually dropped onto your computer when you visit a website. They can serve a variety of functions, but in this particular case Google uses them to collect its tracking data and send them back to UA and GA.
We have gone into the 4 different types of cookies that classic Google Analytics uses in a previous post (which you might want to re-read since it also highlights common tracking problems when using cookies), and the table is shown again below:
[table id=3 /]
On the other hand, Universal Analytics only uses one cookie called _ga and it expires two years after the date it was last refreshed.
Technically, it doesn’t even need the cookie to collect data since it uses the Measurement Protocol. Through this new protocol, UA has the ability to collect data from any device that can send it a HTTP request via the Internet.
A HTTP (Hypertext Transfer Protocol) request occurs whenever your computer or browser fetches a file from a web server.
One of the biggest implications of this is that Google is able to collect usage data from many other devices now – smartphone apps, video game consoles, Blu-Ray players or even cars. Under the previous ga.js, the proverbial cookie crumbles the moment someone logs off her browser and does a conversion action offline. Now, you can potentially send that data (events or hits) back to UA’s servers.
Feature Differences between Google Analytics & Universal Analytics
The screenshot above provides a snapshot of the major differences between Google Analytics and Universal Analytics. As of time of writing, most of the ‘Coming Soon’ features have already been implemented.
So, is Universal Analytics Better?
With Universal Analytics, most of the changes that need to be implemented can be done through the interface itself, rather than hacking the tracking code. This simplifies things greatly and means that more time can be spent on data analysis rather than ensuring that your data capture is implemented properly.