Web Analytics combines multiple Micro disciplines under a huge performance umbrella of Analytics – But what exactly is Segmentation within this area?
Segmentation is one of the most important Possibilities within the Analytics-Universe and therefore one of the disciplines that is worked on the most intense and with the highest development power within Analytics vendors-space.
Segmentation has no limits
There are almost unlimited Segmentation possibilities, meaning combination and filtering of separate metrics or attributes. Before approaching those capabilities in detail, it seems to be important to think about a definition of Segmentation. There are already a lot of diverse sources about a definition (for example on this blog: Segmentation Definition by Thomas Koschel, Head of Analytics at Immobilienscout24 Web Analytics Blog by Thomas Koschel).
In my point of view:
Segmentation describes the act of separating a digital overall picture in single attributes and the adaptation of one or more of those sub-groups for the use of a more detailed examination.
So, segmentation is a daily support activity for the analysis of current or future marketing actions. Especially with the segmentation of Visitors (New Visitors, Returning Visitors, First Buyers, Repeated Buyers) it is a very useful possibility to work on details and execute special communication and/or acquisition arrangements.
Visitor vs Visits vs Channels
Often, the segmentation in visitor or visits will be used: During a visitor review every visits will be attributed to that visitor. On the other hand the visit review is only checking this a single visit for the segmentation purposes.. Furthermore, often it is segmented to each channel: Entry channels, conversion channels etc.
Segmentation is well established in the Web Analytics world as a necessary and important feature. Therefore, it made me all the more filled with disbelief when I saw the following graphic:
This summary was written by a blogger who builds his knowledge on the content of a report from eConsultancy (Conversion Rate Optimization Report by e-consultancy). Once we have learned how important and simple segmentation is to achieve desired results, the positioning appears just above the average “value” line and also operates in the area of “unused” attribute yet. According to this report, A / B Testing and customer journey analysis are the most widely used and most valuable disciplines.
I need to ask you two questions:
- Are A/B Testing and customer journey so valuable because they are new and so many analysts deal with it, which in turn leads to an exploitation of the benefits of these disciplines?
- Why is segmentation so little used, although it a) was available much earlier in the web analytics industry and b) is implemented much simpler than A / B Testing?