Google Analytics segments allow the digital marketer to isolate and compare traffic in a nearly infinite number of ways. Here is a closer look at some of the segments we find ourselves using on a daily basis at Room 214.
Comparing Mobile, Desktop, and Tablet Behavior
You know what’s marginally useful? The Device Category report inside Google Analytics. It’s great if you want to know how many people visited the site on a desktop, mobile or tablet device during a given period. It’s not very useful for anything else.
If you really want to dig into how users interact with your site based on device used, you’re going to need to set up segments for each device category. Mobile and Tablet are system (default) segments but you’ll need to create your own for Desktop. Or import this simple Desktop Traffic segment.
One common use for these segments would be to analyze conversion rate by device. In the screenshot below, we can see that users on mobile devices are much less likely to convert than those on a desktop computer.
Removing “Log In” Traffic
This one is common for SAAS companies. You have a portal for customers to log in to your product. Many of them go directly to the portal for access. Others Google your brand name, click on the top organic result and proceed to click the “log in” link on your homepage. This common workflow can seriously alter your Google Analytics data. You’ll be seeing an inflated bounce rate, low time on site and increased visits from Direct and Organic Search.
What to do?
Filtering out these visits at the view level isn’t really possible, so we’ll have to use a segment. But first: How do we identify these users who are just hitting the homepage in order to log in? We’ll have a to create a Custom Event that triggers when someone clicks the log-in link.
This can be accomplished in a number of ways, but the easiest would to be use Google Tag Manager to set up the custom event. It is also possible to hard code the custom event into the link click. Don’t fret if this isn’t in your skill set, there are plenty of guides available.
Once you have your custom event set up and tracking, you’ll need to create a simple segment that looks like this:
In this scenario we’ve used the Category of the event to create the segment. As you can see, about 15 percent of the traffic to this site is users logging in to the platform
Here comes the hard part. Remembering to enable the segment every time you access the account and making sure everyone else who is accessing the data does as well.
Comparing two geographical regions or eliminating a particular locale can be very useful. Let’s start with the obvious use case: You own business with multiple locations in different cities and you want to know how online behavior differs between regions. That’s easy — set up segments for each location. Google Analytics allows for segmentation by country, region (state) and city.
We can also use this type of geographic analysis to manage budgets for advertising programs. We often compare conversion rates between two locations in order to determine how aggressive our bidding should be in region A vs. region B.
Filtering Out Job Seekers
A common issue we see, especially with smaller businesses, is highly variable traffic from the Referral channel grouping that is directly correlated with the amount of open job listings. Sites like Indeed and SimplyHired will start sending hundreds of visits, skewing your data. There are two ways to temporarily scrub this data for a more pure “marketing view.”
- Identify the websites responsible for the job-seeking referral traffic and create a segment that excludes any traffic from those domains.
- If the external job listings link to a specific page on your site, you could filter out visits that land on that page or pages. This is most commonly an individual page (ex: /careers/data-analyst) but you could also filter out your entire careers directory depending on your site architecture and comfort level with making broad changes.
You can also combine these methods using “And” within the filter for more precise removal or “Or” to cast a broader net.
- User came from a job search site AND landed on a job listing page on your site
- User came from a job search site OR landed on a job listing page on your site
Here’s what the “OR” option looks like:
New vs. Returning and Visit Count
These two segments aim to help you understand how visitor behavior changes over the course of multiple visits. First, we’ll simply look at brand new visits vs returning ones.
The trickiest step here is knowing to look for “User Type” as the dimension. Here is what they look like for New and Returning Visitors.
What if “Returning Visitor” simply isn’t enough? Maybe you’re running an e-commerce operation and want to understand how purchasing behavior differs between a visitor’s first few visits and their fourth and fifth?
You can use the “Count of Sessions” dimension to organize visitors by the number of times a unique user has visited the site. Here’s what it looks like with a segment aimed at capturing visitors with a specific number of visits.
Another method would be to set up segments for a group of visit counts. It might look something like this: 1-2 Visits, 3-5 Visits, 6+ Visits. Here’s what the 3-5 segment looks like:
One thing to consider is that new and returning visitor data is based on cookies. Cookies do not follow users who switch devices or browsers, so take this data with a grain of salt.
That’s not all folks! Here are a few more segments that we’ve found useful over the years.
- Visitors Who Have Converted (Include Goal Completions > 0)
- eCommerce Whales (Include Revenue > $$$)
- Business Hours Comparison (Include Hour Less Than X and Greater Than X)
- Segment by Age and/or Gender
- Organic Image Traffic (Include Referral Traffic matching RegEx “imagedetail|images|mgres”
- Highly Engaged Traffic (Include Session Duration > ?:?? – depending on site average)
- Hostname comparison (Include Hostname = subdomain.domain.com)
There are so many more ways to use custom segments to understand the data housed in your Google Analytics account. While some of these are universal, many of them will be custom built for the problem at hand. Do you have a favorite Google Analytics segment that you keep coming back to? Let us know in the comments.