Join Corey Koberg for an in-depth discussion in this video Segment visitors: User segments, part of Advanced Google Analytics.
- View Offline
- [Voiceover] When advanced segments were first created, they were 100% dedicated to sessions. In a previous video we saw how filters or page views are hit based and segments are session based. And that made it easy to say, "Show me which pages were viewed "in the sessions where a purchase was made." But now, Google's made them even more powerful and given us the ability to look at multiple session as a single group which is tied to that single user. And, in fact, many of you will likely see this as a default option in the interface. So recall in the last video, we focused on the conditions area in the advanced segment session.
Let's head over there. Come down here. Create a new segment. And we went down here to conditions, and we looked at something like revenue. So let's actually pull that up. So here we have revenue and we have this per session or per hit. And most of the time, we're thinking about per session. That's saying the entire transaction no matter how many things we put in our cart. When we finally checked out of this session, how much did we have. Per hit is more about the individual line items. If we wanted to look and see is any one item. But for the most part, we're thinking about this from a session point of view.
Now in that example, we said we wanted revenue to be let's say greater than $100. So the big difference now is we can come up here to this filter and instead of sessions, we can include users. And so now this particular condition has to apply per user not per session. Now here we're still looking at revenue per session although as soon as we change this to users, we get a new option down here which is per user. So now we have the ability to look at revenue per user regardless of the number of sessions.
And what this means is this condition can be met over the course of several visits that may even span multiple dates. You may come back once and then come back a week later and it's gonna think about that and meet this condition as a user not as a session. So both of those sessions will count towards this $100. Now we still have the ability to go down and say "We want to look at a user and then per hit." But that's not really where the main power from this comes. Really being able to do this from a user segment level is a really big deal. This type of functionality is what distinguishes an enterprise tool and Google making this available in the free version is really a great thing for the industry.
At first this was going to be one of the Google Analytics Premium features only, but I'm really glad that they extended this down to everyone in the free version as well. So in case this isn't entirely clear, let's take a little bit more of an example of this. We've talked before about airlines and how they use segmentation for things like frequent fliers as they're easily tiered into a silver, a gold, a platinum, etc. But perhaps we can let the data help us segment this out a little better as well. Airlines almost always sell out at Christmas and other major holidays, so perhaps they're not necessarily so excited about that one traveler who bought that one expensive ticket per year to make that Christmas trip home to grandma's house.
They may want to target the person who buys a medium price ticket every single week. Especially some of those weeks where they're not necessarily filling the plane and they may actually end up with more revenue over the course of the year than that one person who bought the one big ticket. So looking at the data, can we find that traveler who's gonna fly out on Monday and back home on a Friday and do that week in and week out, and try and get more of those folks. Now when you think about this analysis on a simple kind of a grid here. We're going to see multiple visits across the top and then the users on the side.
Now we got the fliers on the side here and the top are the different visits that they took. We saw the first case here, flier number one, who bought the expensive ticket on visit number two. Then he never bought another ticket any other time. So the total amount that they spent was just the price of that one big ticket. Now granted, if we were doing the session type analytics we were looking at before, this flier, number one, would without a doubt appear to be the best. This was the big spender and the person we were supposedly going to look for in terms of our analysis of big spenders. But in the scenario we just laid out, this isn't actually the person that the business case wants.
The business case wanted the second one here. That $400 every single week, day in day out. Now we've got a third one down here who actually spent more per ticket. We see both of these spent more than the top person even though these were the cheapest, lowest fairs there were. And of course, flier number one, the one who bought the most expensive ticket actually ended up with the lowest overall total contribution of revenue. So we can see in this case in our analysis, it's pretty simple here. The fliers you want to target are not actually the ones who are spending a lot in a single session, but rather the most over time.
And what we want to do is have our web analytics reflect the fact that this is the analysis and the business case that we want to look for. So let's go back to the interface and look for a couple different ways to do this analysis. The first thing that I want to do is point out the date range. We can go back 90 days in our user analysis. So this is known as a look-back window and we have a 90-day look-back window in which we're gonna look at the total behavior of a single visitor. We may have data and analytics of a single visitor stretching back over a two-year period, but this user segment is only going to look back 90 days. So the first thing we said we were gonna do is just look by revenue.
So in this case, it's just like we did with the sessions except for it's gonna be users. We select the revenue and we have this greater than or equal to 100. In this case, greater than 100. We can name this user segment USERS over $100. So that's simple enough. But the second way we could look at this is not just by looking at revenue, but we'd also talked about the idea of someone coming back and making multiple purchases. So perhaps our analysis is less about that dollar figure and more about that behavior itself.
So let's do one that shows all the visitors who have over ten transactions regardless of the number of visits. So I just changed this to be transactions. Again per user, per user and we want that to be over ten. Now my data set, I don't have any users who qualify for that. In this one, we don't have a lot of frequent purchasers. In fact, we have virtually no folks who even are coming over two, but hopefully your data set has a few more, and if not this is pretty telling in and of itself being able to do this type of analysis right here.
And of course we want to make sure we don't forget to update our title. Now, I would say that it's also interesting to look the other way. If we were going back and let's say our data set supported the ten one here, we might want to look at people who are doing a lot. But we also want to go the other way. Maybe it's less than three transactions. We want to identify who are these folks, where are we getting them from, are we paying a lot in terms of our marketing to find folks with less than three transactions. So we could do the same thing here. Just flip that around.
In this case, let's say less than or equal to three transactions. Now with these segments defined, we could actually apply them both at the same time. And that would give us even more insight. We could do all the usual analysis we have looked at in the previous videos about which channels they came in from, which marketing efforts were working, which content they can consume, which path they took to the site, and so on. All of those things can be applied, and we can do it for both of those who are in the low value, less than three, those who're above 10, those who are in the over $100 per user. We could look at all those simultaneously. Remember, this is not just about that revenue.
We use this as our first example, but it's far from the only type of analysis you can do. This can really be powerful when we start examining the entire purchase funnel. Particularly if we've done a good job of defining micro conversions of steps along the way. So we could look at people who read a white paper and then they use a configuration tool and then that lead them to purchase something. And this can get even more powerful if we look at the entire scope of multiple visits, but also the sequence in which they happen. Now as you probably guessed, we're gonna save that sequencing for the next video, but for now, I encourage you to spend some time thinking about how your analysis is gonna change and how you think about a segment in terms of the entire user not the just that one particular session.
Learn how to set your account up for multiple users, structure properties and views, and configure the back end of Google Analytics—including a popular WordPress CMS plugin for automating tracking code on WordPress pages. Corey shows how to apply advanced filters and segments (and customize them to fit your needs) and import custom data sets. He also covers debugging and troubleshooting techniques that allow you to fix Google Analytics issues from under the hood, and he reviews the major segmentation and analysis pitfalls to avoid.
Note: This course picks up where Google Analytics Essential Training leaves off. If you have any questions about how Google Analytics works straight out of the box, see those tutorials for more information.
- Setting up a Google Analytics account for multiple users
- Working with the WordPress CMS plugin
- Using custom and advanced filters
- Debugging and troubleshooting Google Analytics
- Using the advanced Tag Assistant
- Setting up advanced segments: custom, user, and sequenced segments
- Avoiding segmentation pitfalls
- Generating better reports and insights from Google Analytics
- Importing custom data