Segmentations

Introduction

Segmentations are the core of Exponea analyses that allow you to group customers into segments based on their attributes and actions they did in your business. With segmentations you can see how different groups of customers differ from each other and then you can use these segmentations in other analyses to get into more detail.

Uses

You should use segmentations whenever you want to group customers based on any metric, for example on the revenue they generated, on their engagement or their activity in your business. You can also group customers based on which funnels they have completed.

Apart from visualization of segmentations, you can also see movements between segments in time, calculate aggregated data such as total revenue for different segments or target customers through scenarios. You can also use every saved segmentation in other analyses, which is usually done to compare the performance of different groups, e.g. from A/B test.

Functionality description

Customers in segmentations are grouped into segments based on rules that you define.

Suppose you want to compare how much revenue came from casual and how much from hardcore customers. You would create 3 segments (for casual, mediocore and hardcore customers) by defining each segment by filters (for example by customer lifetime value). Then you would aggregate purchases for all these segments in a customer metric. Segmentations will then calculate the number of customers in these segments and corresponding revenue share in real-time from all your data.

Rules for segmentation evaluation

  • Customer always belong to the first segment (from left) what conditions match
  • Customer matches empty “Events steps” filter (funnel)
  • Customer doesn’t have to belong to any segment

Creating a basic segmentation

Segmentations creation

Create segments

Use the plus sign button on the right to add new segments. Copy button creates new segment with filters from the last segment. Each segment is defined by set of conditions. See filters documentation for more details.

Each customer belongs to the first segment what match the conditions!

segmentations

Date filters

Date filter can change date for which is the segmentation created. Customers change segments over time and this serves to see movements. You can observe change in segmentation movements panel. Just specify start and end date (both can be in past).

Profile metrics

Basic segmentation shows number of customers in given segment. Add profile metrics to observe other values not only simple counts.

Click on metrics and assign them to X, Y to create 2D visualisation of segmentation – segmentation map. Assign size to 3rd dimension and bubble will have different sizes.

Segmentations profile metrics

Click on segment in segmentation map or segment distribution charts offer following actions:

  • Show customers in segment
  • Send campaign to given segment
  • Use for prediction (experimental)

Name of segmentation is very important. After you save a segmentation a new customer attribute called after segmentation is created. Every day 0:00 CET start process of automatic re-segmentation and attribute values are updated. You can force this process anytime by click on play button in list of segmentations.

Movements

Customer attributes, mainly some aggregates change over time, which causes customers to move from one segment to another. There is a feature in Exponea segmentations that provides also information about these movements, which is shown as a table From->To. When you select a date in the segmentation, this will be taken as a baseline period. Exponea will then calculate how many customers were in the defined segments in previous period, and also provides prediction on the movements to the next period based on the trend.

segmentations movements

Other notes

  • Segmentation creates customer property named after segmentation
  • Every customer what belong to some segment will have actual segment name in this customer property
  • Segments are refreshed daily (around midnight CET)
  • Segment refresh can executed immediately forced with “play” button in segmentation
  • Customer properties aren’t persistent (are stored only in in-memory storage)