How To Use Google Analytics Segments to Analyze Event Data
How to utilize your events in custom segments within Google Analytics. Let’s look at two examples.
We’re really digging in here to some exciting ways to analyze all the event data that you’re pushing into Google Analytics. We just talked about custom metrics and how you can use that to really expand how fast you can analyze your data in bulk. Now we’re going to look at how to utilize your events and custom segments within Google Analytics. I’m going to show two different examples here. Let’s say that you have your product page tagged. We’ve looked at the size guide click as an example as one event, here we’ll see Nordstrom, there’s a size guide. This opens up into a new PDF page. And you want to understand what’s the difference in behavior of people that are interacting with a size guide versus not interacting with a size guide.
We can take this a little bit further. Let’s say there is some discussion amongst your team that you think the size guides need to be redone or they need to be moved on the page, made more prominent, less prominent, et cetera. By tagging this and having this data available in Google Analytics, we can now actually apply a segment, so we can create two segments. A segment of users that view a product page and click on the size guide and compare that to a segment of users who have viewed the product page but do not click on a size guide.
One caveat here, obviously, I’m not showing you actual data from Nordstrom, so take that with a grain of salt. This isn’t real data but it is a real setup that it should make sense as we go through the analysis process. Just quickly looking at the size fit guide click, where we have the events of size fit guide click, this is triggering anytime someone clicks on a size fit guide. In this hypothetical example, anytime someone actually clicks on this.
Inside of Google analytics, you’ll see there are two segments that I have created up top here. And let’s just go down into our eCommerce view. If you have never created a segment before or a custom segment within Google Analytics, basically what you’re able to do is, instead of viewing all of your data, your eCommerce data, your channel data, page view data, et cetera, instead of viewing that through a lens of all visitors, segments allow you… It’s like looking through a different pane of glass. If you’re looking through one window, that is the window of people who clicked on the size guide. And then another window you’re looking through, you’re seeing people that have viewed a product page that have not clicked on the size guide.
It’s really easy to go through and create this. We’ll click on our edit, we’ll see our custom segment, and this would be the same setup if you were creating it from scratch, and you’ll just click on this plus icon. We have our size guide click, which is what we’re naming the segment. And you’ll have all these options here for us in the conditions. The advanced conditions, we are stating here the event action matches regex size fit guide click. When you go through and create this segment, you’ll actually get a little summary over here to the right, just to make sure that you are setting this filter up appropriately. Let’s say I were to add just a bunch of fluff at the end of this. We should get zero percent. This is the way, when you’re going through and creating it, this will validate that you have the segment at least set up in a way that you’re going to get data back from it.
If we look at this other segment, we’ll discard our changes, if we look at our other segment, this is basically doing the reverse. We want to look at pages that have products in the URL. In this case, on Shopify store, you’re going to have products in the URL. Then we want to exclude the event action. We want to make sure that anybody, we don’t want to include anyone in the segment that clicked on the size fit guide. This segment here is people who clicked on a size guide. This one is people who did not, but they were ultimately looking at a product page.
Once these segments are created, then inside of analytics, you’ll start to see you have two different rows of data. We have an example here where users who are clicking on the size guide on product pages, they are converting at nine percent versus people who are not clicking on the size guide. They’re viewing product pages are converting at 0.87%. That’s a pretty stark difference. And if we are just evaluating this one feature and you could take this and come up with ways to test it or make it more prominent or whatever it might be. That’s a pretty cool way of taking data, applying it to GA and really putting that advanced analysis on it by creating two different segments to compare.
Example number two is, let’s take Macy’s, I’ve just mocked up an example here where we have a main navigation click of shop by category or shop by designer. Let’s say there was discussion of either getting rid of one of these or moving the order around or potentially adding additional categories into the main navigation besides these two. This setup is really largely the same. We have event action created. That’s tracking both of these. We have our two segments here, so let’s just click on the edit. We’ll see the conditions, event action matches regex shop by category click. And then our opposite here, we have our event action matches shop by designer click.
Now what we’re doing is we are evaluating users who clicked on shop by category versus those who clicked on shop by designer click, and we can see the conversion rate differences. This one is not quite as widespread as we saw with the size fit guide click. What you can actually do is you can start navigating through your other reports. Let’s jump over here to our channel report. Now we can see by channel how many people have clicked on the shop by category or shop by designer and see the percentages for these, and ultimately the conversion rate for these. If there is a pretty stark difference here, like we have with email, where the shop by category click has a lower number of users that are initiating that event, but the conversion rate is almost 50% higher than those who click on a shop by designer click. You might create a strategy of, okay if we see our email audiences really performing better on they’re shop by category click, so we’re going to move that a little bit higher and have that more prominently located in either a mobile menu or the desktop menu.
You can go through all the different reports here, whether it’s by your source medium, by campaign. This is a really powerful way to start thinking about personalization of your site, so if there are certain features or enhancements on your site where it’s performing better on either a device type or a channel that you can start personalizing that. Whether you’re using Google Optimize or Dynamic Yield, Optimizely, etc., you can use that to start to drive some of your strategies for personalizing based on really what your users are reacting to.
Those are two examples of taking your custom events. Let’s start back at Google Tag Manager. We have our Google Tag Manager event of just a simple size fit guide click. We looked at how to attach a custom metric to this so we can see all of our pages and ranking and see what pages are driving the most size fit guide clicks. Then we’ve taken it one step further of evaluating that event action against one another, so people who have taken that action versus not taken that action and viewing the conversion rate differences for that. And then the last step is actually taking action on it. Implementing your strategy for leveraging a particular interaction that’s converting 10 times more than those who do not take that action.
Take that, think about your events starting back from the beginning of what you want to start tracking on your site, start to think about the segment analysis of how could you analyze different behavior. You can see the step-by-step process that we’ve gone through, and very powerful, very fun stuff. I could talk through all these different ways to slice and dice event actions and behavior for hours on end. Please let me know if you have any ideas or questions of your own, because there’s no end or limit to what you can do and what you can push in the GA to analyze this.