Segmentation: How do I explore a Segmentation change over time using Fixed Time Points?

Segmentation change over time can be examined using Fixed Time Points or Elapsed Time Points.

The default Fixed Time Points option allows you to examine the number of records in each segment at a specific point in time and can be useful for identifying, for example, recently lapsed customers or your current "best" customers.

Where it can be useful to track segments that can potentially change very quickly - for example, post campaign interactions on a website or in store - you can also report on less than a day intervals using DateTime variables. For a worked example of this see Sub-day Fixed Time Segmentation.

 

Segmentation Selections

With this function, each selection used must include a date element that will give the potential for the result to change as time progresses. For example, the following date rule will identify dates between a year ago up to yesterday. Therefore this will result in a different set of dates every day that it is run.

  • Open a new Segmentation tool and drag on the selections you wish to examine

In the example below we have 4 selections identifying High Frequency High Value (HFHV) customers, Low Frequency High Value (LFHV) customers etc.

You can recreate these selections as follows:

1 - HFHV – Booking Date variable set to The Last Year. Apply RFV to show frequency as >=2 and value to show Average (Mean) Cost to be >800

2 - LFHV – Booking Date variable set to The Last Year. Apply RFV to show frequency as <2 and value to show Average (Mean) Cost to be >800

3 - HFLV – Booking Date variable set to The Last Year. Apply RFV to show frequency as >=2 and value to show Average (Mean) Cost to be <=800

4 - LFLV – Booking Date variable set to The Last Year. Apply RFV to show frequency as <2 and value to show Average (Mean) Cost to be <=800

We will also create a selection to show people who have not made any Bookings in the last 12 months. This group can be added to the display later when we look at the migration patterns between periods.

5 – No Bookings – Booking Date variable set to The Last Year. Apply RFV to show frequency as 0 (N.B. no distinction has been made here between previous purchasers and new customers)

Set the Reporting Points

To determine which points in time you wish to report on, select the Reporting Points icon on the tool bar:

By default the display will be set from today back to the start of the year 3 years ago and be a Fixed Date Schedule - i.e. using a fixed From and To date. It is also possible to create fixed time segmentations using relative date rules by selecting Date Rule Schedule from the Type drop-down in the middle of the schedule panel.

Adjust the options within the window that displays:

In the above example the movement between the selections will be shown as at the 1st January on each of the relevant years.

The more reporting points you set, the longer the processing time to see the results

  • Click on the Time Report tab and then click the Build button

This will produce a cube that displays the number of customers in each segment for each of the reporting points.

  • Click on the Chart tab on the right hand side of the window to view the movement of numbers graphically

When we look at the graph above we can see all 4 segments have an upward trend in 2015. However in 2016 we see a downward trend apart from the LFLV which still continues upward (not a great trend!).

Before we examine the migration between periods more closely we should firstly include the segment that defines no bookings have been made. We need to do this so that we can see new/returning customers and those who are lapsing.

  • Drag the 5 - No Bookings variable into the Segments tab

To view the migration of customers between our segments within 2 specific points e.g. 01/01/2016 and 01/01/2017

  • Click on the Migration (2 points) tab and enter the From and To dates and then click the Build button

In the above example we can see that of the 40,185 people in HFHV on 01/01/2016, 83 are still in HFHV on 01/01/2017, but 206 are now in LFHV, 196 in HFLV, 425 in LFLV and 39,275 made no bookings at all.

To take this a step further we may wish to see the movement between all the points we set on the Reporting Points tab.

  • Click on the Migration (All points) tab and click the Build button

In the above screen shot let us examine the 4th row down 1 – HFHV -> 4 – LFLV. We could say this shows the movement from our best customer group to our worst customer group. Between 01/01/2015 and 01/01/2016 772 people made this move, 425 in the following period and then 1,404 in the period after that.

In this example it looks like we have high movements between these groups with the last figure being particularly high. However, we might get a clearer picture if we change the display to show Percentage of Segment at Start Point.

We can now see that same trend in the proportion of people in the HFHV group in each period moving to the LFLV group.

The more troubling figures in our example are the migrations to the No Bookings segment.

The highlighted rows in the above screen shot show a very high percentage of customers in each segment not making a booking in that year. This would indicate our organisation has a serious issue with customer retention.

 

To create and review Time Reports using Elapsed Time Points, see Segmentation: How do explore a Segmentation change over time using Elapsed Time Points?

For information on the use of aggregated expressions with this tool, see Segmentation: Using Aggregated Expressions

 

Back to Segmentation Overview