This is the third post in a 3 part series on Google Analytics setup Best Practices. Google Analytics is something we sometimes take for granted, but in reality sometimes accounts with years of data have to be abandoned simply because they were not set up properly. Hopefully this series will help people understand some of the best practices for setting up Google Analytics.

Once you have your accounts, profiles, and user access accounts set up for your Google Analytics account, it's time to look at creating filters.

Filters are actually a really powerful tool which you can use to manipulate the data coming into your account. Think about that for a second.

How Google Analytics Collects Data

Let's take a couple of minutes to look at how Google Analytics collects and records data, and hopefully filters will make a lot more sense. I'm taking some artistic license here for simplicity, but this is the process in a nutshell.

Basically, you drop the javascript snippet on the website. A tracking cookie is set on the visitors browser. A single pixel image is also requested from Google, and included with the image request is all the information about the visitor. On a very basic level, the image call looks something like this:

When the tracking gif is requested from Google, the information about the visit is collected in Google's logs. Every hour or so, the logs are accessed, and the information is pulled through your filters and added to your GA database. Once the data is in the database, it (currently) can't be changed. What gets recorded, stays recorded.

What this means is that filters are a way of changing the data you collect about your visitors, before it gets recorded in the database. You can say "only collect this data", or "don't collect that data". When you think about it, this makes filters a really powerful tool.

Google Analytics Filters

Filter Overload

There are a few filters that people use often. Some of the most popular filters are:

  • Filters that include or exclude your staff
  • Filters that include or exclude only PPC traffic
  • Filters that include or exclude only Organic traffic
  • Filters that include or exclude subdomain traffic

These filters can be very useful in getting the information you need out of the website, and many GA accounts have at least an "exclude my staff" filter applied to their account, and that's great.What gets a bit unwieldy, however, is when multiple filters are not managed with a strategy in mind.

Frequently, I've seen accounts which have several profiles, and each profile has the same filters replicated over and over in each profile. So for example, if you have an "exclude staff" filter, you create it in profile #1. Then you create profile #2 and think "I have to exclude my staff", so you create a new filter for profile #2, and so on.

The problem with this approach is that it recreates filters over and over, and you end up with something that resembles a tangled mess. If you need to update the IP address of your "exclude staff" filter, you have to update it in every single staff filter for every single profile. This is not only time consuming, but it's also prone to human error missing a profile or pasting in the new filter information.

Bad Google Analytics Filter Architecture

The better way is to create one filter to do the job, and use it in all your profiles. Instead of creating a new filter for each profile, choose the "Apply existing Filter to Profile", and choose the existing filter that you need. Once you apply the filter, you can also use the "Assign Filter Order" tool to choose which filter should be applied in which order.

Google Analytics Filters

Once you have one filter for what you need, and you have added the single filter to each profile in your account, you should now have a much more streamlined Filter architecture, like so:

Good Google Analytics Filter Architecture

How to Test New Filters

As we discussed last time, it's always a great idea to test any new filters in a new profile, to make sure that it does what you think it does. Only once you are sure that it's doing what you want, then port the filter into the profiles you need.

If you're in a technical mood and have some time on your hands, you can also read the full explanation on how Google Analytics works.

Read the whole series:
#1 How to Set Up Your Google Analytics Account Structure to Avoid Headaches Later
#2 Google Analytics Profile Architecture for Dummies
#3 Google Analytics Filter Best Practices