Segment containers in Adobe Analytics

Hi. I’m Ben Gaines from Adobe Analytics product management. And, in this video I want to talk about the different containers that you can use when building segments in Adobe Analytics. Containers are really important concepts when you’re segmenting your data, because they’re going to control the scope or the way that the data is queried so that you have better control over what actually comes back and you can make sure that you’re segmenting your data the way that you intend to segment your data. There are three types of containers that Adobe Analytics offers, and the easiest way to see them all is right here in this show dropdown in the segment builder. The first one and the most granular one is “Hit”, and then you have “Visit”, and then you have “Visitor”. And, to see the impact of changing the container, we’re going to use the segment preview here in the upper right. And, that’ll let us see how the data that we’re getting back, the data set that Adobe Analytics is actually returning to us, changes as we change the container. And, hopefully that’ll help us master the way that we use containers and segments. So, as I mentioned, the most granular container type is “Hit”, and to demonstrate this I’m going to use a very simple segment defined as “Page equals home”. Now, when you have the hit container applied, the way to kind of read this is that the data set with this segment applied is going to be comprised of all hits where the page name equals “Home”. In other words, all hits to the home page. Now, “Hit” can mean page view, it can mean “link click”, really it’s a server call, or anytime that data is being sent into Adobe Analytics. In this case, since we’re talking about page name, it’s going to essentially mean page views. So, my data set that returns is going to be all of the page views where my page name equals “home”, or you know, all views of my home page. You can see that reflected here in the preview. I’ve got about, in the last 90 days, I’ve had about two point six million page views on my site, and about one hundred and forty-three thousand of those were on the home page, which seems about right, intuitively. And we can also, just because you’re using a hit container, doesn’t mean that the visits and unique visitors metrics aren’t valid. What this is saying is that there were ninety-seven, almost ninety-seven-and-a-half thousand visits that had a hit on the home page, and about eighty-one thousand unique visitors that that saw the home page during that time period. Now, where this becomes really relevant is that anything else that happened during those visits and unique visitors during those visits, or for those unique visitors, won’t be included in this segment, because I have “Hit” selected as my container. So, if I had a visitor who viewed the home page and then viewed the product page, any of the data that was collected on the product page will not be included in this segment. None of that data will be included in this segment, because again we’re only looking at the page views that occurred on the home page, and we get back from that request all of the data that was collected on the home page. So, I can still run my referring domains reports or use the campaign variable in a project, and those will work, but only for data that was captured on the home page. Now, let’s move up to the visit container. And, when I do that, I want you to watch the segment preview, because it’s going to change pretty significantly.

So, the number of visits increased only slightly the number of unique visitors. I think very, very slightly increased, but the number of page views increased dramatically. And the reason for that is that when, you say “visit”, you’re now including all of the data from any visit that saw the home page. Whereas, with “hit”, you were looking at data collected from just that home page. Now, you’re looking at everything that happened before the home page was viewed and after the home page was viewed, during any visit where the home page was viewed. So, for example, if a user clicked on a campaign that I’m running and landed on a different landing page, not my home page, and then they viewed a few product pages and then they finally clicked over to my home page, all of that data would be included in this segment, and that’s where the page view number increases dramatically, because now all of those other page views that occurred during the visit are included in this segment and are returned and in the segment. Often, users of Adobe Analytics who are newer to segmentation will build a segment using the visit container to do something like this, where they have a page in there, and then they’ll run a pages report or use the page dimension in a project and wonder why there are all of these other pages showing up, when clearly they specified in the segment that they wanted where page equals “Home”. Well, now you all know the reason why. It’s because all of those other page views are included. Any visit that saw the home page in this case, or whatever other dimension you choose would be included in here, as well. Now, let’s look lastly at the visitor container.

And again, watch the preview when I make the change. What you should see is the number of page views should increase a little bit. The number of visits should increase. And, let’s see if that’s what happens. Yep. Exactly what I expected. The reason is very similar to what we saw in the previous example, going from “Hit” to “Visit”. The visitor container includes all of the data from any visitor who had any visit that ever saw the home page in this case. So, let’s say you have a visitor who during the last 90 days, which is our date range here, the last 90 days has made five visits to my site, and they only viewed the home page during one of those visits, maybe the second visit, and on all the other visits they went straight to product pages or other landing pages. All of that visitor’s data, all five visits, are going to be included in this segment. Not just the visit where they saw the home page. That’s what the visit container is for. But all of that visitor’s data. So this is really useful, doing the visitor container is really useful for doing analysis over a long period of time, where maybe you want to see visitors who responded to a certain campaign. What did they do over the next five months? And so, you can use a visitor container and say visitors who saw a certain campaign and then apply that over a long period of time, and you get all of the data for those visitors so that you can see things, behaviors that you wouldn’t be able to see if you were only looking, for example, at the visit where they saw the campaign in question or the page in question. So, really, as I mentioned at the beginning, it’s all about the scope of the data and the question that you’re trying to answer. Sometimes you want to do visit level analysis. Sometimes you want to see all of the data for a page, for a certain dimension or a certain metric, and you only want to see data that was captured at that same time and that’s where “Hit” would come in, “Visit” in the previous example, where you’re doing visit-based analysis where you want to see, for example, certain paths and then visitor for longer term analysis where you want all of the data for any visitor who matched that segment. This is definitely something that takes a little bit of getting used to, but, once you get the hang of it you’ll be a segment ninja and you’ll be able to slice and dice your data with great ease and do the analysis.