Mirror of Oracle documentation

Converted for search and offline reading. Authoritative source: Oracle. Diagrams and some complex tables are simplified — check the PDF when in doubt.

5 Customer Segmentation

This chapter describes the Customer Segmentation module.

Introduction

Customer segmentation is an enterprise-specific solution that uses data mining to group customers based on customer attributes and customer transactions. The retailer can use this information to describe and predict customer behavior. It provides the retailer with a vehicle to target customers with offers, pricing, assortment, and experience.

Retailers understand that shoppers are heterogeneous in nature, that they possess different wants and needs, and that it is impossible to satisfy them all. Retailers can differentiate themselves from their competitors by specializing and offering goods and services that are tailored to one or more market segments.

Customer Segmentation can be used to group customers and to discover hidden customer segments based on the contents of customer shopping baskets and the number of shopping trips they make. Loyalty card data is used to determine if these customer segments differ in terms of socio-demographic or lifestyle characteristics and whether these characteristics can be used to target different customer segments with more relevant product offers.

Retailers can create localized assortments and use customer insights to determine which products to offer by location or channel. This provides insights into the importance of a product to key customer segments and helps when making drop or keep decisions about products within an assortment.

This science-driven tool helps to automate segmentation in a repeatable process and bridges the gap between targeted marketing to targeted assortments.

Features

The key features of Customer Segmentation include:

  • Scenario-based segment generation, based on customer attributes, customer behavior, and transactions.

  • Attribute importance and correlation mining to identify significant attributes and their associations.

  • The ability to generate granular customer segments via the UI. These customer segments include departments such as men, women, and children, health and beauty, or groceries.

  • A three-step segment-generation process.

  • What-if capabilities that can be used to create multiple segmentation scenarios and then measure them against one another. This can help ensure that the most appropriate segments are used by the applicable planning and execution processes.

  • Automatic ranking of segment scenarios to support what-if comparisons.

  • Recommendations for the optimal segment scenario and number of segments.

Table 5-1 Cluster Criteria Overview Tab

FieldDescription
NameThe criteria ID and user-assigned name of the segment.
Segment ByThe Segment By option used for the segment.
Created ByThe name of the user who created the segment.
Last Updated ByThe name of the user who most recently updated the segment.
Last Updated OnThe date when the segment criteria were most recently updated.
StatusThe most recent, up-to-date status across the scenarios for the
segment criteria. Value include Created, Ready for Preview, Ready for
Approval, Approved, and Rejected.
Period CountThe number of calendar nodes defined for the criteria. Hover over the
count in order to see a list of the calendar keys associated with the
criteria.
Merchandise CountThe number of merchandise nodes defined for the criteria. Hover over
the count in order to see a list of the merchandise keys associated with
the criteria.
Location CountThe number of location nodes defined for the criteria. Hover over the
count in order to see a list of the location keys associated with the
criteria.

Segmentation Criteria

The following segmentation criteria are supported by default:

Customer Demographics

This descriptive segmentation technique leverages customer loyalty programs and demographic information (such as residence, profession, age, gender, ethnicity, marital status, and education) about customers to generate demographics based on customer segments.

RFM and Customer Behavior

Segmentation based on purchase behavior aims at discovering groups of customers who exhibit similar purchasing behavior. However, the definition of behavior in this context includes many factors. For example, retailers may want to distinguish between light and heavy users, regular stock-up shoppers versus emergency top-up shoppers, lunchtime shoppers versus evening shoppers, home and daytime shoppers versus work and weekend shoppers, or fastcheckout customers versus regular checkout customers. The two important behavioral dimensions for understanding customer motivations are visit behavior (identified by the time of day and the day of week that the visits take place) and shopping behavior (identified by the customer’s spend dispersion across categories purchased during the trip).

Category Purchase Behavior

Another type of behavior segmentation aims at segmenting the customers who seek similar benefits when evaluating and choosing or purchasing products. These benefits can be measures such as economical price, bulk products, durability, or free shipping. Here, the segment process considers factors that capture customer sensitivity to price and promotions for each category. This can help retailers to segment customers by distinguishing predicted customer responses to the targeted or general promotion of products.

  • Execute version. This is available when the setup is complete.

  • Create a customer segment using an existing version. This opens the Generate Customer Segment tab. The New Segment Criteria pop-up is displayed, and the default values are filled in. (See the defaults that are selected on selection of the version in segment criteria.) This option is only available if the version has a status of Ready for Use.

Table 5-2 Version Details

NameDescription
NameThe name of the version.
Criteria CountThe number of criteria associated with the version. When you click the
link, a pop-up is displayed that lists the segment criteria details that are
associated with the version.
Last Updated ByThe name of the user who updated the version most recently.
Last Updated OnThe date when the version was last updated.
Created ByThe name of the user who created the version.
StatusThe current status of the version. Values include Version Setup
Complete, Filtering Completed, Sampling Completed Successfully,
Attribute Mining Completed Successfully, Version Ready for Use.
MerchandiseThe merchandise node defined for the version.
LocationThe location node defined for the version.
System Generated VersionThis column provides a flag indicating whether or not the version has
been created using a batch process and is set up with a default
configuration.

Version Criteria Pop-up

This section describes the Version Criteria pop-up, shown in Figure 5-4.

i——CSsSCSCSC‘éA

override these top categories further while defining versions for merchandise. By default, the top categories in the selected merchandise are driven based on the sales share for each category. You can remove categories to reduce processing time and storage, using the user interface and adding emerging to the top category list.

Version Status

The version status has one of the following values.

Table 5-4 Version Status

StatusDescription
Version Setup CompleteIndicates that version setup is complete.
Filtering Completed with ErrorsIndicates that version filtering has completed with errors.
Filtering Execution in ProgressIndicates that version filtering is in progress.
Filtering Completed
Successfully
Indicates that version filtering has completed successfully.
Sampling Completed with
Errors
Indicates that sampling has completed with errors.
Sampling Execution in
Progress
Indicates that sampling execution is in progress.
Sampling Completed
Successfully
Indicates that sampling has completed successfully.
Attribute Mining Completed
with Errors
Indicates that attribute mining has completed with errors.
Attribute Mining Execution in
Progress
Indicates that attribute mining execution is in progress.
Attribute Mining Completed
Successfully
Indicates that attribute mining has completed successfully.
Version Ready for UseIndicates that all three execution phases (filtering, sampling, and
attribute mining) have completed successfully. Once the version is
ready for use, it is ready for the creation of segment criteria.

Generate Customer Segments Tab

The Generate Customer Segments tab is used to create segments and then model the segments with various scenarios in order to determine the best set of segments. It consists of three stages: Segment Criteria, Segment Results, and Segment Insights.

Actions» View [9§ YA * Actions» View> Fomaty & 2B) ® ABR Tt wW i ae & Fa

All Segment Criteria Filters j Merchandise 9-CHAIN ow | Location o-CHIN w | Calendar Winter Jan 16 ¥

ee

eee reer ener reer ee eee ener

eee

ay

Table 5-5 (Cont.) Pop-Up Details

Field NameDescription
Scenario ExecutedThe number of scenarios executed for the segment.

All Segment Criteria Scenario List

This displays the scenarios for the selected segment criteria in the All Segment Criteria tree.

Table 5-6 Scenario List

Field NameDescription
NameThe name assigned to each scenario that has been created for the
segment.
StatusCreated, Ready for Preview, Ready for Approval, Completed with
Errors, Approved, Rejected.
RankThe system-calculated rank for the segment.
Optimal # of SegmentsThe system-calculated optimal number of segment centers.
User PreferredIndicates whether or not the user prefers the segment.
System PreferredIndicates whether or not the system prefers the segment.
# of AttributesThe number of attributes that were used in the segment.
Max. # of SegmentsA user-provided value for the maximum segment centers that the
segmenting process should consider.
Min. # of SegmentsA user-provided value for the minimum segments centers that the
segmenting process should consider.

Segment Criteria

In this pop-up, you define the initial segmenting parameters for the segment criteria of a new segment. Note that the hierarchy type supported can be configured at the time of deployment.

Figure 5-10 illustrates how to use a simple approach to segmenting by selecting attributes from a Segment by. For example, you can select the RFM and Customer Behavior Segment by and generate segments using the total number of trips and the amount spent by customer.

Table 5-8 Effective Period

OptionDescription
Fiscal PeriodIf you select this option, choose the period and the subdivisions of that
period from the drop-down lists.
Planning PeriodSelect from the range of values provided for the period. Planning
periods are user-defined buying periods for a season or a season
subset.
Select DatesIf you select this option, choose the start and end dates using the
calendar pop-up.

Summarization

Data summarization is available by default and set to either Category or Sub Category. It is applied to Category Purchase Driven Segment by. The segmentation process considers the top selected categories and their attributes and groups customers based on their sales patterns.

Source Time Period

Source time periods are selected based on the version selected for the segment criteria.

Table 5-9 Source Time Period

FieldDescription
Period LevelSelect from Fiscal Year, Fiscal Quarter, Fiscal Period, or Fiscal Week.
Start PeriodOnce you select the Period Level, you select the starting subdivision within that
period.
End PeriodOnce you select the Period Level, you select the ending subdivision within that
period.

Contextual Area

When you are creating new segment criteria, you can see details about the following parameters that can help you understand the segment you are creating.

Segment By Hierarchy

The following information is displayed when you select a template or use the icon to select the Segment by.

Table 5-10 Template Display

PropertyDescription
Template NameName of template configured during deployment.
DescriptionDescription of template.

Table 5-10 (Cont.) Template Display

PropertyDescription
Segment ByA predefined group of attributes that include Customer Demographics,
RFM and Customer Behavior, Category Purchase Driven. These
criteria types are sets of attributes. For example, customer
demographics are the properties of a customer. These properties can
include ethnicity, income, and age.
Segment By Primary Scenario

You see this when you select Segment by in the Criteria panel when you are setting the segment criteria parameters or when you select Segment by in the contextual area for the hierarchy.

The system displays the primary scenario, its preconfigured properties, and the significant attributes identified during the attribute importance process for each segment by.

The following information is displayed.

Table 5-11 Primary Scenario

PropertyDescription
NameThe name of the primary scenario.
StatusCreated, Ready for Preview, Ready for Approval, Completed with
Errors, Approved, Rejected.
Maximum # segmentsThe maximum number of segments. The default value is 20. This is
used for analyzing the segments.
Minimum # segmentsThe minimum number of segments. The default value is 1. This is used
for analyzing the segments.
AttributeA list of the attributes configured during segmentation.
Attribute importanceThe weighted average across attributes and importance index for each
attribute.
Planning Period

This list displays the time period you selected for the segment criteria. This information is available only for planning periods, where it provides the start and end dates of the planning period. This content changes whenever the planning period is selected in Effective Period when you are setting segment criteria parameters.

Explore Data

Use the Explore Data pop-up to examine data for the segment criteria you defined. You can view the customer and attribute summary that provides input into the segmentation process.

Process

In this pop-up you can view criteria and attribute summaries as well as their related contextual BIs.

Summary

The Summary lists the criteria you initially selected to define the segment.

Table 5-12 Explore Data: Summary

FieldDescription
NameThe name you provided for the segment in the Segment Criteria stage.
Segment ByA predefined group of attributes that include Customer Demographics,
RFM and Customer Behavior, Category Purchase Driven. These
criteria types are sets of attributes. For example, customer
demographics are the properties of a customer. These properties can
include ethnicity, income, and age.
MerchandiseThe merchandise level and nodes for the segment.
LocationThe location level and nodes for the segment.
Fiscal PeriodThe time period for the segment.
Merchandise Hierarchy TypeDetails about which type of hierarchy the segment criteria have been
created for.

Attribute Mining

This screen provides you with insights about the attribute mining process, which lists the attributes’ significance and their correlations. The attributes required for the customer segmentation process come from different dimensions such as customers, their households, demographics, and purchasing behaviors. This process helps to eliminate redundant attributes and to identify the attributes that may have the most influence on generating customer segment. The attribute summary displays information about data availability and data quality by providing the attributes’ distinct values, percentage of nulls, and statistics summaries, such as mean, median, and standard deviation.

Attribute Importance

Along with attribute summaries, the system generates an attribute importance index that sums up data quality, data distribution, and its representation of each attribute in the data. See Figure 5-11.

Filters |: Customer Attributes All

w © | Products All

Correlation Matrix

The correlation matrix displays attributes, attributes values (in the case of discrete attributes), and correlations between customer and products. Each cell in the matrix provides a visual indication of how attributes are correlated, along with the correlation value. Various colors indicate the strength of the correlation among attributes, with -1 and 1 indicating strong association.

Table 5-15 shows the strength of the correlation for the different ranges of values for the correlation coefficient.

Table 5-15 Correlation Coefficient Values

Correlation Coefficient ValueStrength of Relationship
1.0 to 0.5Strong
0.3 to 0.5Moderate
0.1 to 0.3Weak
-0.1 to 0.1None or very weak
-0.3 to -0.1Weak
-0.5 to -0.3Moderate
-1.0 to -0.5Strong

Contextual Area

This area provides graphical illustrations of the detailed data distribution about the customers and their attribute importance.

Analyze Customers

In Explore Data, the BI displays the data distribution of the customers by each participating attribute as well as other configured informational attributes. Customer Segmentation identifies the bins based on the underlying sample data and displays the histograms. It provides the percentage of customers that are present in a selected location. For example, a company may have ten percent of premium customers who are high spenders and who shop frequently.

|

Table 5-17 Attributes

Field NameDescription
ParticipatingA check in this column indicates that the attribute participates in the
segment criteria.
Attribute GroupA logical grouping of attributes such as demographics or purchase
behavior.
AttributesAttributes that are potential candidates for generating segments.
ImportanceSystem-generated attribute importance index that indicates the
significance of each attribute.

The Attributes toolbar includes the following functionality:

Figure 5-18 Attribute Toolbar

Table 5-18 Attribute Toolbar

FunctionDescription
Action menuResets the attribute selection to the default selection that system
identified using attribute importance thresholds.
Include or exclude attributesAny attribute beyond a certain threshold is not included in the
segmentation process.
Contextual Area

The contextual business intelligence lists a set of attributes that the current scenario includes as the participating attributes for the segmentation process.

|

Table 5-25 Scenario Compare

Field NameDescription
Max. # of SegmentsThe value used for the maximum in the scenario execution, if this
option used.
Min. # of SegmentsThe value used for the minimum in the scenario execution, if this option
used.
Optimal # of SegmentsThe value used for the optimal number of segments in the scenario
execution, if this option used.
RankThe value for the rank.
Is System PreferredIndicates whether the scenario is the one the application prefers.
Is User PreferredIndicates whether the scenario is the one the user prefers.
Smallest Segment SizeThe size of the smallest segment.
Largest Segment SizeThe size of the largest segment.
Has OutlierIndicates a segment with the number of customers below a threshold.
For example, the number of customers is below a certain percentage of
the number of customers in a segment.
AttributesA list of relevant attributes.

Scenario System Recommendations

The application provides the following recommendations at the scenario (segment set), segment, and customer levels.

Scenario Optimality

This graph indicates how the system identifies the best number of segments for a given data set. It starts with a small number of segment centers and searches for the number beyond which there is minimal dispersion. At this point, increasing the number of segment centers any more only reduces dispersion by a small amount, and the marginal improvement is small.


In this guide