I’m a self confessed Google Analytics (GA) addict, I drive everyone a little crazy when I get going on how much you can do with it. In a perfect world I would have historic GA data on any website before touching it in any way.
If you’re considering redevelopment you can analyse which pages are most and least popular, which are turning people away, how many people are viewing your site via a mobile device and so much more. If you’re trying to optimise for search you can gauge what phrases are working, where people land on the site and what actions they complete – even down to how much money they spend on what items right back to the keyword they put in. So… very… cool.
Out of the box Google Analytics is good; with a few tweaks it’s fantastic. Here are some fairly simple modifications you can make to the basic set-up in order to get more accurate reports from the system.
The most basic installation of GA requires you to sign up for an Account and insert a small block of code (your Tracking ID) into your website template. This code is used to track visitor activity on your site through the use of browser cookies and send it back to Google who compile it and provide it back to you through the form of a dashboard.
It’s really important to ensure you set your reporting time zone accurately during the sign-up process because it can’t be changed later. I would also recommend setting up a separate Account per website if you have multiple websites because it keeps things so much neater.
Once you have GA installed on your site you can already start to get a good picture of visitor activity on the site and have taken a step in the right direction. However at this point there are a few more things you can do to improve data quality. The sooner you do this the better because it is not possible to back-date or rerun your statistics once they’ve been compiled.
Profiles are the foundation to all of my suggestions below – they allow you to have multiple reports from the raw data provided by your tracking code and apply specific rules to each. Individual profiles can be shared with other users – you don’t have to share log in details and can ensure that people only see data relevant to them.
You’ll find Profiles on a tab within the Admin area of GA. When you set up a new Tracking ID the system creates your first profile and names it ‘All Web Site Data’. You can add your own by clicking on ‘+ New Profile’, then choosing a name and a time zone.
I recommend adding a new Profile called ‘Web Site Data’ and setting this as the Default Profile (under Property Settings). Any appropriate filters and rules should be applied to this profile leaving ‘All Web Site Data’ without any rules applied. This ensures that you can see exactly how the data comes in naturally and provides a vital back up if anything is set up incorrectly.
If you want to play with some Filters or Goals it might be worth setting up a test profile especially for this purpose and moving them across to the main profile when you’re satisfied they work as desired.
Filters allow you to manipulate the data coming into GA before it’s compiled into your reports. There are many potential uses of this but I’ve covered the ones I find most useful below – traffic exclusions and avoiding duplication.
Filters are found in the ‘Admin’ > ‘Profiles’ section of GA. Ensure you select the correct Profile by clicking on it’s name before browsing to the ‘Filters’ tab, here you can view any filters which are already set up and add new ones.
Set up Traffic Exclusions
My first port of call is to set up filters which exclude both clients and my visits from the reports. On low traffic sites a large percentage of visits will come from the site owner and their developers; this data can really skew statistics such as new vs. returning visitors, average time on site and average number of pages viewed.
To set up traffic exclusions you should first find out the IP addresses of any locations you wish to exclude. The most client friendly way to get an IP address is by asking them to use a site such as www.whatismyip.com from each location and note down the resulting addresses.
You then need to be within the Profile section as mentioned above, click on ‘Add Filter’, give it a name which will make sense later (e.g Home IP Exclusion) and choose ‘Pre-defined filter’ > ‘Exclude traffic from the IP addresses’ > ‘Equal to’ and then plumb in the IP address of the relevant location and press save. Repeat for each location that you wish to exclude and then from that point forward GA will ignore any access from devices using those internet connections.
Avoid Duplication caused by Case Sensitivity
My second use of filters is to set up modifications to ensure that key data reports into GA as non-case sensitive. By default GA would count ‘Blue Widgets’ and ‘blue widgets’ as two separate entities and effectively create duplicate entries in your reports. Most would count these as the same thing so you can reduce confusion through the use of filters.
There are a host of variables where you might choose to do this – the main ones being pretty much anything relating to how the visitor landed on the site – Source (e.g. google), Medium (e.g. search) , Keyword (e.g. ‘blue widgets’), Request URI (the address your visitor came from) and Target URL (the page on your site they landed on).
My preference is to create a series of filters which set all these values to lower case before being compiled. To do this you need to be within the correct Profile section as mentioned above, click on ‘Add Filter’, give it a name which will make sense later (e.g. ‘LCase Source’) and choose ‘Custom filter’ > ‘Lowercase’ > ‘Filter Field’ > Campaign Source and press save. Repeat for each relevant field and then from that point forward GA will automatically convert all instances of that field to lowercase before including them in your dashboard.
Coming Up in Part 2
Hopefully this has given you an idea of what Google Analytics tracks by default and helped to make your reports a little more accurate. Part 2 will cover reporting conversions through Goal and eCommerce tracking.