Customer Insight

Customer Insight
Functional Features

Automatic and Intelligent Customer Analysis

Provide the customer analysis capability in wizard mode based on the customer life cycle and automatically identify typical customer characteristics and potential customer segments.

Comprehensive Customer Profiles

Based on customer profiles in the traditional telecom domain, analyze customers' mobile Internet behavior on the social, content, and business websites and in apps, analyze customer behavior characteristics and service preferences based on the customer behavior and knowledge bases, and create 360-degree customer profiles.

Real-Time Customer Profile Updating and Sorting

Access customer behavior data in real time, analyze customer long-term preference by day or month, analyze customer real-time preferences by event, and effectively support real-time and personalized marketing and recommendation.

Function View
Function Description
  • Customer Knowledge ManagementCollapse
    • OOTB Customer Tag
    • The Customer Insight provides abundant preconfigured telco tags and Internet tags to comprehensively describe customer features and generate 360-degree customer profiles.

    • Tag Life Cycle Management
    • In the life cycle of a tag, after the Create, Refresh, Reuse, Disable, Modify, or Offline operation, the tag may enter any of the following states: Draft, Add Approval, Applied, Expired, Modifying, Modify Approval, Offline Approval and Offline.

      Create:

      Configure basic information and rules of a tag and submit the tag for review. An operator who has the review permission reviews the created tag to ensure tag quality. The tag enters the Applied state after being approved.

      Refresh:

      Tag refresh can be automatically scheduled and manually triggered. After the refresh, the system generates the latest user list.

      Reuse:

      Rules of expired, disabled, and offline tags can be reused for quick tag creation. For tags created by importing a file, the Reuse button is unavailable.

      Disable:

      Tags that are outdated or inefficient can be disabled. Then the generation rules or related analysis models of the tags can be optimized. Tags can be disabled in this condition. Disabled tags are not displayed on the tag analysis page. Users who have the restart permission or tag creators can check disabled tags under My Space > To-Do > Disabled list.

      Modify:

      Disabled tags can be modified. Some basic information cannot be modified, for example, the tag name and type.

      Offline:

      Tags that are outdated, incorrect, or inefficient can be brought offline. Offline tags cannot be reused, but their rules can. Offline tags can be viewed only by their creators.

    • Tag Store
    • Manages tag life cycle, including tag classification, tag creation, tag rule configuration, tag version management, and life cycle management.

      Tag category management

      Tags are managed by category. Categories are displayed and used in a tree structure.

      Tag permission control

      Tags are classified into Visible to all and Visible to area. A user who has the All Tag Visible permission can uniformly manage tags and customer segments of all cities.

      Tag version management

      Historical tag version information can be viewed, including the version number, modifier, modification time, and rule information. Comparison between historical tag versions is supported.

      Tag evaluation

      After tags are brought online, users can recognize popular tags, evaluate tag usage effects, and effectively manage tags through the tag evaluation function. The comprehensive score of a tag is specified by the popularity KPI on the tag details page. Top popular tags are displayed on Knowledge Management > Home.

      Tag rule configuration

      The following configurations are supported for creating tags through rules:

      − Supports 1:N customer data

      The 1:1 customer data indicates that a customer can have only one record in one statistical period, and the 1:N customer data indicates that a customer can have multiple records in one statistical period.

      - Support Internet data for tag rule configuration

      The system provides access times-based filtering through processing of Internet access data including the URL, app, and search keyword during tag rule configuration.

      − Support data from multiple periods

      Attributes, tags, and customer segments from different periods can be used for customer segmentation rule configuration. Data on an absolute date, on a relative date, or in the latest period can be used.

      - Support calculation of attributes of the numeric type.

      Multi-period data calculation results (such as the sum, maximum value, minimum value, average value, growth rate, function calculation, comparison with the average value on the whole network, comparison with the historical average value, comparison with the average value of a recent day) of attributes of the numeric type can be used.

      - Support usage of tags as rule conditions.

      During tag rule configuration, tags can be used as conditions and calculations can be performed on the tags.

    • Customer Profile Capability
    • Customer profiles are detailed customer information queried by existing attributes and tags, which helps better understand customers and identify customers' requirements.

      In the current version, the location analysis capability is supported. The system ingests original location CDRs, calculates the data through the time and space noise suppression module and stay and moving point calculation module, aggregates the data by time and location to generate location data, and provides the location data for location-based customer segmentation.

    • Customer Segment Analysis
    • The Customer Insight suite provides the interactive customer segmentation capabilities, including:

      Customer segment management by category

      Manages customer segments by category. Categories are displayed and used in a tree structure.

      Customer segment permission control

      Customer segments are classified into public and private customer segments. A public customer segment can be used by all users. A private customer segment can only be used by users of the department or area to which the customer segment creator belongs.

      Customer segment data period

      Tag data integrity included in customer segment can be viewed. Tag rule factor data can be viewed on the tag details page.

      Customer segment version management

      Historical customer segment version information can be viewed, including the version number, modifier, modification time, and rule information. Comparison between historical customer segment versions is supported.

      Customer segment download

      Customer segment list download audit is supported. The customer segment list file to be downloaded can be encrypted.

      Location-based segmentation and distribution visualization

      Location attributes can be selected for specifying time segments and location information during customer segmentation rule configuration.

      Customer segment rule configuration

      − Supports 1:N customer data

      During customer segmentation rule configuration. The 1:1 customer data indicates that a customer can have only one record in one statistical period, and the 1:N customer data indicates that a customer can have multiple records in one statistical period.

      − Support data from multiple periods

      Attributes, tags, and customer segments from different periods can be used for customer segmentation rule configuration. Data on an absolute date, on a relative date, or in the latest period can be used.

      − Support calculation of attributes of the numeric type.

      Multi-period data calculation results (such as the sum, maximum value, minimum value, average value, growth rate, function calculation, comparison with the average value on the whole network, comparison with the historical average value, comparison with the average value of a recent day) of attributes of the numeric type can be used.

      - Support usage of tags as rule conditions.

      During customer segmentation rule configuration, tags can be used as conditions and calculations can be performed on the tags.

      Customer Segment Data Analysis

      Customer segments can be analyzed for displaying customer distribution information. The following analysis capabilities are supported:

      - Trend Analysis

      - Composition Analysis

      − Comparative analysis

      - Primary Factor Analysis

  • Customer Knowledge DiscoveryOpen
    • Real-Time Marking
    • Real-time marking is a process of marking users' network access log files entered by the CAE in real time and adding details such as users' network access behaviors, accessed objects, accessed apps, and accessed web pages, which provides data for event processing by the CAE. Real-time marking has high requirements for real-time processing and needs to be implemented in the streaming processing platform.

    • Offline Marking
    • Offline marking analyzes and marks users' Internet access logs, and recognizes information such the app, web page, behavior, behavior object, and browser. Offering marking has low requirements for real-time processing and can be centrally implemented in a regular Hadoop cluster.

    • Managing the Feature Library
    • Feature library management provides unified management of feature files, fields, and categories.

      Add:creates a feature file and adds feature fields to the feature file. New feature fields can be added to a feature file only after the feature file is disabled.

      Enable/Disable: enables or disables a feature file or a field in the feature file. If a feature file has enabled feature fields, the feature file cannot be disabled and can be disabled only after these fields are disabled. Feature fields can be enabled and disabled in batches.

      Modify: modifies basic information about a feature file or feature field.

      Delete: deletes feature fields in a feature file. Feature file definition cannot be deleted. Feature fields can be deleted in batches.

      Import: imports metadata package and data file.

    • Configuring Composite Rules
    • Offline-marked detailed data of different dimensions can be combined to generate composite rules. Tags of specific composite scenarios can be obtained through composite rules.

      Composite rules can be defined by time, location, app, web page, Internet behavior, and behavior object based on the site requirements.

    • Composite Tag Library
    • After composite rules are configured, the system automatically generates detailed tag information based on the composite rules. You can search the tag library for detailed composite tag information.

    • Knowledge Base Management
    • Knowledge Base Rule Configuration

      The Corpus provides preconfigured rule knowledge bases. The rule knowledge base files need to be modified or imported again after being updated.

      Knowledge Base Search

      You can search knowledge bases to obtain detailed information about Internet resources such as apps, websites, terminals, videos, and music.

      Metadata Management

      Knowledge base metadata describes knowledge bases. The knowledge base metadata of the system includes the following information:

      − Knowledge base name and file. The knowledge base names and types, and storage directories of knowledge base files are stored.

      − Dimension information. Directory levels of knowledge bases or category information are stored.

      − Dimension value information. Dimension value information of different levels in the knowledge bases is stored.