Segmentation is the core of analytics as we all want to try to understand and predict human behavior. Segmentation analysis will be better, faster and more accurate if we use data science. Adobe Analytics uses data science methods to evaluate your segments under Segment IQ. When we compare segments without statistical tests we really can't be sure that the perceived differences are for real. It might be just a coincidence.
Adobe Analytics Segment IQ have 2 modules.
Segment comparison panel
Comparing segments in fallout
Segment comparison panel:
Adobe Analytics Segment Comparison is one of the data science modules that Adobe offers. You can access to Adobe Analytics Segment Comparison under Workspace Analysis. Segment comparison is one of the Panels you can add to Free Form Panel. You will drag two segments into the panel, and you are going to view couple reports and visualizations that shows statistically difference between two segments.
All you need to do is to choose 2 segments and hit build. Adobe Analytics will run its data science process at the background and will create several visualizations and reports. Adobe's data science module compares every single metric, dimension and segment you have. You will see that process is really fast, if you had to do this analysis manually, you have to spend couple days or more.
On the left side you will see how your segments are different for specific metrics, segments and dimensions. These 3 reports will hopefully give you insights about your segments.
Top metrics against segments
Top dimension against segments
Top segments against segments
Here we want to check column called "difference score". A difference score of 1 means it is statistically significant, while a difference score of 0 means there is no statistical significance. Adobe will show results that is close to 1. Then you can see real values of each segment under their columns.
If you want to evaluate certain segments, metrics and dimensions you can do that by choosing those metrics in the advanced options before hitting analyze button.
But this tool will be useful if you really collect very different and unique data. If you are collecting very standard data, you might not get much.
Top metrics against segments
Under this report we see if any of our metrics are showing different behavior under these 2 segments. Here Launches/Visitors has a difference score of 1 which means the difference we see is statistically significant. This means people who came from paid search launch the app more (0.42) versus people who comes from Natural Search (0.37). Although difference is not huge, it is still different. Here I am using dummy data, so if you see something weird please ignore and check your own data to see if tit makes sense. Here on the row 2 Lifetime value is zero for both with a difference score of 1, this is due to dummy data I use. Please ignore:) And don't forget one of the most important quality of an analytics intelligence is to to find the truth within sea of data by cleaning the irrelevant data which can happen a lot with digital data.
On the right there is a trend visualization of both segments for the first metric here but you can also create trend visualizations for each metrics here by hovering over next to the title.
Top dimensions against segments
This report is similar to the previous one. Instead of metrics we look at dimensions. In this example I want to show you one issue with this report. It is not a real issue but something we can't avoid. As you can see here the first values that show up here is not a new information for me, here of course I will see campaign data as the top variables. Because we have chosen every metrics, dimension and segment here. an example where we need to exclude some values. We can exclude some values at the beginning or we can remove them here with hovering over the value and hit delete. Whenever you see see variable that does not help you in this table, just remove it.
After I removed the information I already know, here I start to see other variables. And as you can see here we could not observe any significant difference for these segments. And do not forget and conclude it certainly: this does not mean that these segments have no difference, this segment do not have difference based on segments I collect under this Adobe Report Suite.
Top segments against segments
This report also works with the same method, here our objective is segment. Again here with my dummy data this is all I have, not much.
Default report create some default visualizations and if you want you can change and add different visualizations. In order to show overlaps between 2 segments, you can also use Venn diagram. This visualization will show you how certain value overlaps for 2 segments. Now it is time to go and use the report and if you have any questions ask me in the comments.
If you want to get most out of your Adobe Analytics implementation you can contact me on Linkedin for consulting or training. We both provide implementation services and business analysis and training services. I have 14 years experience to implement and use Adobe Analytics with big brands like Expedia, American Airlines, Best Western International, ING Bank, Vodafone, Buhler Group. Unfortunately many Adobe Analytics experts out there only tool analysts, they do not have significant product development experience. I have been developing digital products since 1999. Adobe Analytics has become my biggest support for an insight and that's why I learned it very well. I am certified Adobe Analytics architect.
My other passion is food, visit my site to see all my recipes:)
