Hey everybody, it’s Doug. In addition to answering analytics questions that you know you have, one of the cool things about analysis workspace in Adobe Analytics is the ability to answer questions that maybe you didn’t know you had. You know, when you don’t know what you don’t know and this is where anomaly detection comes in. Anomaly detection uses AI to show you where your expected numbers would be in time series data like we have here, and then to show you when your actual data falls outside of those expectations either in a good way or in a not so good way. Now you can enable anomaly detection just in a table as you can see here, or you can also show it in a line graph that I have above. Let’s take a look at the table as I scroll down here on my visits and page views. And I scroll down and you see I have these triangles here on these specific days for this number. And as I mouse over that and it has this little square and I click on that and it’ll say an anomaly was detected on that day for this number, it is 72% above expected. Well that’s great and I can do the same thing on these other fields. So I can go over here and I can click on this one and go, it was 28% below expected. So we can see some of these outlier numbers on these specific days. And if I scroll down I think there might be a couple more down there, yeah, well one more down there. Now if you, especially if you had a lot of these numbers that might fall outside of the expected range, it might be hard to kind of click through them and really digest this. So that’s where that line graph comes in handy. So I’m going to go back up. We’re going to click back up here to the top so we can see both our visits and page views. And you’ll see that we have these little round dots in there. And if I mouse over those, it will show me that that means there was an anomaly detected and this one was 17 below expected and this one was 72% above expected. And you can see that these different dots are representations of either being below or above the expected numbers. Now, as you can also see as you mouse over that it will show you that range, kind of that band of, you know here’s kind of an expected range of numbers that wouldn’t be too anomalous. And it’s also a little easier if I kind of get rid of either page views or visits. Let me just click on page views. That’ll go away. Now I’m just looking at visits and you can see this band of expected data and you know, all this is falling within that expected band until we get to this day. And that was actually, you know, 9% above the expected. So that’s good. You can see this one falls a little bit below. This one is above again, et cetera. And so these numbers will show you if it’s above or below. And again, you can kind of see that in the band. Now, if I turn on page views by clicking on there, turn off visits, then you can also see that band without even having to mouse over that and you can see where that should be and where you’re falling. Now, there is also a dotted line which is kind of exactly where it is expected. Sometimes you can’t see that because it is where the actual numbers fell. But in this case where you can see it definitely wasn’t in that spot what was a kind of a very thin range but that one was 28% below expected. And then you’ll want to analyze that with contribution analysis which we will talk about in another video. But just wanted to show you this anomaly detection that is happening on our data here when it is represented in this time series fashion. Now, if you don’t want it to show up, you don’t want people to see this and get sidetracked by this anomaly detection data. You can go up to the line chart and you can turn it off right here, show anomalies, turn that off, that’s great. Click out of that, and I won’t see that, and I won’t see that for visits either, as you can see there. But if you go down into the table it’s still going to be there. And so if you don’t want these markers to be there either, you can get rid of those per column. So if you go up to each column and go to the column settings then you can get rid of anomalies in each of those columns. And there it is, and you can get rid of those, so that way you can decide whether you want to see those anomalies or not in your data, and be able to see those times when you should dig a little deeper to find out what was happening on those days. It caused either a higher number or a low number. Good luck. -