Analyze Segments

Prev Next

The Analyze Segments report calculates overlaps and indexes for segments and datasets with the signal data, and is used to identify characteristics of interest for audience building as well as further investigation and modelling.

Access the report from Analytics | Analyze Segments:

Note!
AnalyseSegments is considered an INTENSIVE process - it can take some time to execute depending on the size of segmentation data being analysed.   Often AnalyseSegments will be scheduled as part of a load process.  When working with the Analyze Segments report, use the 'sample' option until the report is ready to be run against the full set of data

Required Inputs

For the Analyze Segment report to run, the following inputs are required:

  • Target Dataset(s)
  • A set of Segment IDs

Target Dataset

  • One or more Target Datasets- these can be provided as individual datasets, or as part of a collection.  
    • NOTE: Target Datasets represent the primary groups being analysed.    Typically these are:
      • Profiling fields: Data attributes (e.g., demographics, behaviours) used to characterize populations, such as Age/Generation, Gender, Income, Occupation etc
      • Population(s) of interest: Specific groups, such as Recent Purchasers, selected for comparison.
    • Target Datasets must be ON or LINKED TO the PRIMARY CONTACT TABLE (this is the table that contains the unique Signal IDs.  Example names of the PCT are Entities, DATA_EYE etc)
    • See Target Dataset Selection for an overview of working with Target Datasets
    • The report will calculate overlaps such as, "How many records in each Segment are ALSO in Target Dataset 1".

NOTE
The report will process whichever window (DataSets or Collections) is displayed.   This allows for easy switching between different configurations in the report.



Segment IDs

  • At least 1 Segment ID must be provided for Analyze Segments to run.  One output row will be generated for every segment that is analysed.
  • Segment IDs can be selected in the following ways:
    • Filter: A query that produces a column containing Segment IDs that match the Segment IDs in the Signal database.
    • List: A fixed list of Segment IDs, generated from one of the following:
      • Project Explorer | Datasets | Segments
      • Signals | Segment Selector
      • Analyze Segments | Segment Filter | Open Segment Selector
    • Model: Dynamic segment lists can be generated using the following:
      • Profile: Create a profile of Segment ID from a table that contains a Segment ID column (e.g., the Load History table).
      • Advanced Search:

NOTE
The report will process whichever window (Filter, List or Model) is displayed.  This allows for easy switching between configurations in the report


Calculation Settings

Calculation Modes

The following calculation modes are available:

  • Default / Blank (Count):  Returns a target and overlap column for each provided target dataset
  • Index: Returns a target, overlap and index column for each provided target dataset
  • Index Only: Returns an index column for each provided target dataset
  • Ranking: Ranking has the following modes:
    • default - rank is calculated from [Tgt] column
    • Overlap - rank is calculated from [Ovlp %] column - [Ovlp %] = Ovlp Tgt
    • Target Total - rank is calculated on [Ovlp %] column - [Ovlp %] = target / target total
    • Overlap Total - rank is calculated on [Ovlp %] column - [Ovlp %] = overlap / overlap total
  • Target Non Target
Index Calculations 
See AnalyseSegments for details of how the index is calculated

Ranking Modes

Ranking: sub-mode = default/none

Ranking: sub-mode = overlap


Ranking: sub-mode = target total


Ranking: sub-mode = overlap total


Output

Output to:

  • File - an excel file will be created on the server
  • Model - a data table model will be added to the system and can be accessed via Project Explorer | Reports | Models
  • Table - a table will be created and can be accessed via Project Explorer | Database

Use the "Export to Excel" button on the micro-toolbar to export the results to Excel

See AnalyseSegments for details of output columns and calculation details

Interacting with Results

It is possible to interact with the output of the Analyze Segments report in the following ways:

  • Extract Datasets - select cells from the grid tab and drag to Hotbar
  • Use the scratchpad tab to perform ad-hoc calculations and apply formatting
  • Output current display to Excel
  • Send the JSON for the current report to script

Sample Mode
When working in sample mode, datasets dragged from the grid will NOT match the counts in the grid, as datasets are always calculated against the full database.

Saving Reports

Analyze Segment reports can be saved and reopened.    When the report is reopened it will be in Sample Mode by default.


Note
An opened report will also be in view mode and will display stored results. To access report settings and recalculate, click the settings button.  Make sure the sample setting is as desired before calculating.