Cohort

A Cohort is typically defined as a group of people who share a common characteristic over a certain period of time.

A Cohort report breaks down user data into groups that share common characteristics or actions within a defined time frame. For example, to see how many users registered to your website in the last few weeks broken down by week and then see how many of them returned.

6-34

This example report shows a row (cohort) for each week indicating users who were active (meaning performed any event) in the website. Each column represents a bucket of users –

  • The leftmost column named Users indicates the total number of users who were active during the first week.
  • Column 0 represents the number of users (of those that were active in the first week) that came back and were active in the second week. The second week starts from the date shown in the second row on the left (meaning 2016-05-19).
  • Column 1 represents the number of users that were active in the second week (as represented in Column 1) that came back and were active in the third week and so on. The second week starts from the date shown in the third row on the left (meaning 2016-05-26).

This report can also be displayed as percentages, by turning on the Percentage slider 6-35.

The example above shows that as time goes by, fewer and fewer users are coming to the website (to begin with), as shown in the lower rows of the Users column for each week. The report also shows that as time goes by, the chance of these users returning increases, as shown in the buckets (columns) of the lower rows.

Note: Any event can be included in a Cohort report. However, if you include an event for which the Include in Path option is not selected, then a Where condition must be defined for this event name with an exists operator; otherwise, empty results may be returned.

6-36

To create a new Cohort report:

  1. In the CoolaData Administrator console, click Reports reports.
    – OR –

    In the Dashboard window, click the Add Reports 6-15 button. A list of reports is then displayed.

6-2

  1. Click the + button to add a new report. A dropdown menu of report types is displayed. Select Cohort. The following displays –

6-37

  1. In the Users who field, select the first thing to happen –
    • Did anything – Users who performed any event.
    • Did – Select a specific event from the dropdown menu.
  1. In the And then field, select the next thing to happen –
    • Did anything – Users who performed any event.
    • Did Select a specific event from the dropdown menu.
    • Did not do Select an event that the user did not do from the dropdown menu. This means that the user did nothing more or performed any event excluding this one.
    • Did nothing – No events occurred within the date range of the report.
  1. In the Group cohorts by field, select –
    • Time – Define the timeframe of the cohorts. For example, 2 Days.

6-39

    • Limit – If Group cohorts by equals Time, then the Limit specifies the maximum number cohorts that are displayed, even if more can fit into the time range of the report. The timespan of each cohort is determined by the Cohort every field (described above).
    • Count Users –
        • In each cohort – Each user event is counted in each cohort that it appears.
        • Once – Each user event is counted once – in the first event in which it occurs within the filters and time range of the report.
    • Property – Select a property by which to group the report.

6-40

  1. In the Measure field, select the thing to be counted and displayed in the report –
    • Select a function, such as COUNT, MIN, MAX, SUM, AVG and so on.

6-41

    • Select the thing to be measured, such as Users or a property. When Users is selected, only the COUNT function is available.

6-42

If you selected Users, then an additional field is displayed –

6-43

In each bucket – Counts each user each time that user’s event appears in a bucket (for example each week or day).

Once – Counts each user once, the first time that user’s event appears in a bucket.

Recurring – Only counts users that were active in consecutive buckets, meaning if the user is not active in a specific bucket, then that user is not counted in that bucket and in subsequent buckets even if they are active.

For example, if a user performed an event in bucket week 0, 1 and 3, then –

      • If In each bucket is selected, then that user is counted in weeks 0, 1 and 3.
      • If Once is selected, then that user is counted in week 0.
      • If recurring is selected, then that user is counted in weeks 0 and 1.
  1. In the Bucket every field, define the timespan of each bucket. Limit – Specifies the maximum number buckets that are displayed, even if more can fit into the time range of the report.

6-44

  1. Define the report’s date range and define the report’s filter if you would like it to be different than the dashboard in which it appears.
  2. In the Description field, fill in any description of the report.

6-9-1

  1. Click Apply and then click the Compute button to display an example of the results of this query, as shown below – !!!!!!!!!!!!!!!!

6-45

  1. Select one of the graphic representation icons. !!!!!!!!!!!!!!!

6-46

As you select various visualization options, the chart appears at the bottom of the window accordingly.

    • Table
    • Single number visualization
    • Line chart
    • Area chart
    • Bar chart
    • Column chart
    • Pie chart
    • Geo chart
    • Pivot table

Set the visualization options and sliders that appear underneath according to your preference –

6-47

    • Theme – Selects the color to be applied to the report.
    • Show Bucket 0 – Percentage –Hides the first (leftmost) column, which represents bucket 0.
    • Percentage – Displays percentages in the report instead of quantities.
    • Running Total – Displays a running total that accumulates the columns. This option is only relevant when the Once option is selected in the Measure field.
    • Captions – Enables you to configure various titles that appear in the report.
  1. To publish this report –
    • Click the three dots in the top-right corner of the page to display a toolbar.

6-48

    • Select the Publish option.

6-50

    • Fill out the window to specify the recipients to receive this report, the frequency and the time of day.The new report now appears in the Dashboard and is sent to the specified recipients daily.
  1. Click Save.

 

Cohort analysis breaks data down into groups which usually share common characteristics or actions within a defined time frame. Cohort analysis helps identify patterns in the life cycle of customers, adapt, and tailor the service to specific cohorts. A group of people who share a common characteristic or actions within a defined period form a cohort, represented by a row in the returned result set.

The date range selected in the report will specify at which date the first cohort begins.

Cohort

To define a cohort:

  1. Choose the start and end actions: these can be anything, nothing, or specific events.
  2. Add any conditions to the start and end actions: click the arrow icon to expand the step conditions.
  3. Group cohort by time or property: If grouping by time, users can be counted once or again in every cohort.
  4. Choose what to measure: Other than counting users, you can also measure the value of any other property over time for these users. If counting users they can be counted in one of 3 ways:
    • Once: only counts the user once on the first occurrence of the event throughout his life cycle
    • In each bucket: the user is counted each time he performed the event
    • Recurring: each cohort user is counted only if the user’s activity occurred consecutively in each bucket throughout the duration of the cohort.
  5. Choose the bucket duration: buckets are the collections of users by time when the second step was performed, measured from the time the first step was performed. You can choose to group by any time range.
  6. Define limits for cohorts and buckets: if cohorting by time you can choose how many cohorts to include. If cohorting by property, the cohort will show all available values for this property.

 

Reverse Cohort

It is also possible to analyze a sequence of two events in the reverse order, i.e. what happened prior to a certain event.

The same definitions apply, with the change of the first step happening after the second step.

Reverse Cohort

Print Friendly, PDF & Email