Go to ...

Troubled Projects

Special Interest Group

Troubled Projects on Google+Troubled Projects on YouTubeTroubled Projects on LinkedInTroubled Projects on TumblrTroubled Projects on PinterestRSS Feed

SAS – Enterprise Intelligence – Churn and Marketing campaign Administration Resolution For Telecom Field


In the fashionable Telecommunication with the levels of competition mounting up among the support suppliers, consumer acquisition and retention is a sizeable challenge. For the new entrants, attaining the new shoppers is the highest precedence, while for the incumbents, retaining the income earning shoppers is crucial.

The telecom providers can increase profitability by creating a predictive modeling for determining possible churn candidates and non-income earning shoppers and can increase income and profitability by focused campaigning and marketing presents which will not only retain these shoppers but also change the non-income earning shoppers to profitable income earning shoppers.

This post highlights the necessity of churn and marketing campaign management and the usage of SAS – Telecommunication Intelligence program (TIS) for the reason. It also involves numerous implementation worries for SAS – TIS in the serious time scenario.

Churn Administration

Customer acquisition and retention is a major challenge in all industries. In the Telecom sector it impacts profitability of the organization if a consumer churns ahead of the organization can receive again the investment decision it incurred in attaining the consumer. As a result, it is extremely vital to identify the profitable shoppers and retain them.

With the telecom market place becoming additional competitive, analyzing the reasons of the consumer leaving the support of the organization is increasingly tricky. In this circumstance, it is even additional tricky to predict the probability of the consumer to leave in in the vicinity of upcoming. It is increasingly demanding to devise a price-effect incentive to goal the suitable consumer to influence him to continue to be with the organization.

Predictive modeling of churn assessment and management aims at creating scores depicting the probability of the shoppers to churn out in upcoming. This can take into thought distinctive facets of customer’s susceptibility to churn, like the record of people people who have churned in the past and construct a facts product that generates an effortless-to-recognize reference quantities (scores) assigned to every single shoppers. These shoppers are then focused with incentives to prevent their cancellation. In other text, Churn assessment decides the possible reasons for a upcoming cancellation depending on the past data which will support the providers to customize their offer. For case in point: if assessment reveals that quite a few shoppers have churned from a distinct region last month and further investigation has identified that there are repeated simply call drops (disruptions in support) in that exchange (or BTS region). It can be concluded that due to the technological inadequacy of that distinct exchange, repeated simply call drops are knowledgeable which has contributed to the consumer dissatisfaction and their going out of the organization. So further technological answer for that exchange can avert upcoming possible churns.

Enterprise Definition of Churn Administration

Defining churn is the very first and foremost exercise in Churn Administration planning. Distinctive providers outline churn in accordance to their company activities.

Churn definition differs from a Pre-compensated to Post-compensated scenario.

In pre-compensated scenario, a consumer can be thought of as churned in the next conditions:

a) If the consumer goes out of community (deactivated)

b) If the consumer is an active non person (ANU)

A consumer can be thought of as ANU when:

i. the consumer has no outgoing or incoming usage for last (X) rolling days

ii. the consumer has only incoming usage but no out-likely usage for last (X) rolling days iii. If the customer’s usage is beneath a pre-established (company decided) volume for last (X) rolling days.

In publish-compensated scenario, a consumer pays a rental on monthly foundation. So in case of non-usage or decrease-usage, the organization earns preset income from just about every publish-compensated consumer. As a result, the consumer is thought of as churned only when he/she goes out of community (Deactivated).

Churn Parameters for company assessment

Following defining churn, subsequent exercise is determining the right parameters for the contribution of churn. The churn probability or churn scores for unique shoppers can be created on the foundation of next categorical aspects:

1. Customer demographics Customer demographics linked facts are used for segmenting the whole consumer base depending on:

a) Age

b) Intercourse

c) Profits

d) Customer Account Details

e) Subscription lifestyle cycle

2. Billing and Usage:

Billing and usage linked data which is acquired from swap (Connect with Facts Documents) is predominantly used for detection of churn probability. The next aspects are used:

a. Cost program

b. Regular monthly usage summary (Billed simply call rely, Billed facts quantity, Free of charge simply call & Facts volume)

c. Regular monthly income contribution

d. Bounced payment

e. Running channel data

f. Recharge channel data

g. Network Products data ( Voice, Messaging, Facts)

three. Complex High quality:

High quality of support is a possible churn driver as simply call drops or inferior support good quality improves the consumer dissatisfaction and for that reason churn probability. In case of CDMA, as the consumer is tightly coupled with the handset equipment, the ageing of handset impacts the probability of the consumer churn.

The next aspects are used:

a. Dropped simply call counts

b. Service good quality

c. Devices age (Handset age in case of CDMA)

four. Agreement Particulars: At the close of the contract interval or grace interval, the probability of the consumer leaving the relationship is superior, for that reason it has a superior impression in resolve of churn. The next aspects are used:

a. Commitment interval

b. Count of contract renewal

c. Present-day contract and close date

five. Event linked:

Loyalty scheme or loyalty rewards are critical motorists for retention. The Loyalty scheme linked facts is used for churn scoring.

Pinpointing the supply units:

Following choosing the Churn parameters, subsequent move is to identify the supply units from in which the respective facts will be extracted.

For case in point:

Cusomer aspects from CRM technique

Usage & Billing linked aspects from Billing technique

Complex High quality from Exchange & CellSite

Activation aspects from Provisioning technique

Facts Administration

Facts management is the foundation for a company assessment. Right facts should really be existing in right position.

Facts Administration has 3 areas:

Extraction: Entails extracting of facts from supply technique and loading to facts interchange layer

Transformation: Entails validation of the extracted facts (eg: Validation for special keys), development of signing up for disorders among the the tables, cleaning of invalid facts and many others.

Load: Entails loading the facts in the Enterprise Intelligence Facts Warehouse

Facts Modeling and Churn Score era

When the authenticated facts is available in the facts warehouse, the facts modeling is executed. It is an iterative course of action. The good quality of the product is accessed and the product which returns the greatest company value is thought of. This product provides benefits in the kind of churn rating of unique shoppers which can be used for analyzing marketing campaign targets.

Working with the churn scores for Retention Campaigns

The facts product generates unique customer’s churn rating which ranges from to 1.

– Signifies the very least probability of the consumer to churn

1 – Signifies highest probability of the consumer to churn.

These scores are weighted parts of numerous parameters, this sort of as

Usage data

Stability data

Recharge data

Decrement (Advertising and Main) data

Handset characteristic

Network coverage

High quality of support

Customer support/grievances

Cost program sensitivity

Enterprise selection needs to be taken to decide an higher threshold of the churn rating. The shoppers higher than this threshold need to have to be analyzed further (eg: shoppers with rating .seven and higher than). The top rated two parameters contributing to the churn rating to be created on unique consumer level (for shoppers getting churn scores bigger than the threshold). Dependent on these parameters retention marketing campaign can be carried out. The parameters can be as follows:

Usage figures: The usage behavior can be derived from the combination of decrement (promo and main), equilibrium and recharge data. The consumer who has greater rating in “lesser usage” can be focused with marketing selling price program presents to greatly enhance his/her usage and change that consumer from non-income earning to income earning.

Bigger Off-web usage: The greater rating on “off-web usage” signifies that the distinct consumer has called extremely routinely to other networks. A focused marketing campaign can be executed with the selling price program beneficial to simply call other networks. A further assessment of the called off-web quantities can outcome in determining routinely called off-web quantities which can be focused by strategies as a candidate of acquisition.

Handset Features: The handset used by the consumer can be aged and be missing the fashionable characteristics. In this case, the probability of the consumer to change to a newer handset is superior and there is a sizeable susceptibility of that consumer to transfer to another support company getting bundled handset offer. A retention marketing campaign can be focused (to this team of shoppers getting superior Handset churn rating) with new support offer bundled with handset.

Customer Service/Grievances: The greater rating in Customer support/Grievances signifies that the consumer has called the consumer care routinely and probability of that consumer dissatisfied with the support is greater. Additional investigation to the consumer simply call conversation aspects can reveal the induce of routinely calling to consumer support. Following the execution of strategies on the foundation of the churn rating and churn motorists, the marketing campaign reaction needs to be captured and fed into the database for assessment of successfulness of strategies.

Implementing Churn Administration Resolution Implementation Measures

The next phases are involved in Churn Administration answer implementation:

1. Need Evaluation: In this phase, the company prerequisites are gathered and analyzed and company definitions for churn are decided

2. Resolution Assessment: In this phase, the company intelligence options are assessed with the superior level requirement of the applying organization. The feasibility test is performed depending on the superior level company requirement and facts availability.

three. In depth Evaluation/In depth design and style: In this stage, the company prerequisites for the Churn Administration project are analyzed in depth for design and style, development and enhancement of the project. An exercising is executed to recognize the availability/unavailability of data expected to fulfill the company prerequisites and facts mapping from supply technique.

four. Facts Evaluation – ETL: In this stage, the facts is extracted from the supply technique, transformed (cleaned/modified for lacking fields and facts good quality is analyzed) and then loaded into Facts Warehouse of the company intelligence tool.

five. Facts Modeling: In this stage, the analytical facts styles are produced by statistical solutions (eg: Logistic regression technique) on historical facts for churn rating prediction and Analytical Foundation tables are populated by facts.

6. Reporting: The churn rating (-1: – indicates fewer probability of churn, 1 – Greatest probability of churn) is created at every single consumer/account/membership level and corresponding report is created.

seven. User Acceptance Take a look at and Roll-out: On completion of productive UAT, the program is rolled out for the company people.

Implementation Problems

There are a number of worries when a company intelligence answer is carried out in a substantial scale of thousands and thousands of shoppers.

The important time of the implementation is eaten by facts management. Facts management utilizes 75% of the total implementation time. Facts Administration involves:

Identification of supply units from in which facts needs to be extracted:

Due to the involvement of many supply units (CRM, Provisioning technique, Billing, Mediation units and many others.), it becomes increasingly tricky to identify the right supply technique for numerous facts fields. Identification of the right facts supply and mapping to DIL fields consumes greater part of the implementation time. If the facts supply mapping is erroneous, then the subsequent actions of implementation (modeling, assessment) will also be erroneous. As a result, particular care needs to be taken throughout the facts gathering exercising.

Facts High quality: Facts acquired from the supply units need to have to be of superior good quality and mistake free. The important challenge in applying a company analytics answer is acquiring a superior good quality facts. Cleansing up of facts and filling the lacking fields consume sizeable volume of implementation time.

Adjust management: With the implementation of a BI answer, the people need to have to change the way they used to conduct churn prediction and marketing campaign management. As a result, person adaptability and person consciousness needs to be constructed up by good coaching periods

To make the Enterprise Intelligence technique operational: Following the implementation, specific organizational composition for handling the BI operations needs to be planned and the means need to have to be properly trained in the expected parts.

SAS in company analytics

SAS is a major company analytics program and support company in the company intelligence domain. It has shipped confirmed options to obtain pertinent, reliable, reliable data through the companies assisting them to make the suitable choices and achieve sustainable general performance advancement as properly as mitigate risks.

SAS has an prolonged capability of handling facts of significant scale (with the support of SAS-SPDS – scalable general performance facts server). This mixed with powerful programming language and enriched graphical interface has differentiated it from the other analytical tools available in the market place. This makes SAS flawlessly suited for organization usage in which it calls for handling of substantial facts suppliers.

SAS – Telecommunication Intelligence Resolution (TIS)

SAS has a number of industy specific options. SAS has packaged their company analytics expertise in the kind of styles, procedures, company logic, queries, reviews and analytics.

TIS is the telecom sector specific company analytic answer which has been constructed specific to telecom sector needs. This answer assists the telecom support suppliers with specific modules, for case in point:

SAS Marketing campaign Administration for Telecommunication

SAS Customer segmentation for Telecommunication

SAS Customer retention for Telecommunication

SAS Strategic Effectiveness Administration for Telecommunication

SAS Cross sell and Up sell for Telecommunication

SAS Payment risk for Telecommunication

SAS churn management and marketing campaign management answer involves Segmenting the whole consumer base

Detecting the will cause of churn

Scoring the unique consumer on the foundation of their churn probability

This churn rating is further used as an enter for marketing campaign management.

SAS Facts move (Architecture)

The facts needs to be collected from numerous supply units.

CRM technique: Customer/Account/Subscription linked facts

Provisioning technique: Activation date, equipment (Handset) age Billing Process: Billing facts

Mediation Process: Connect with record aspects

The facts is collected in the Facts Interchange Layer (DIL). The facts is then extracted, transformed and loaded into In depth Facts Shop (DDS).

The facts is used for:

1. Dimensional Facts Modeling: This is used for query, reporting and OLAP (On-line Analytical Processing)

2. ABT (Analytical Foundation Table): This is the answer specific product produced which can be used for a distinct assessment. For case in point: The ABT for churn product.

three. Marketing campaign Facts Mart: This facts is used for focusing on specific consumer segments for focused marketing campaign.


As a result, it is essential that churn management is an crucial challenge in the fashionable working day Indian telecommunication sector. Detecting the good rationale of churn and predicting churn in progress can save the organization from significant income decline.

Enterprise Intelligence tools support the telecom support suppliers to perform facts assessment and to predict churn probability of a distinct consumer. Aside from churn predictive assessment, the tools can be used for numerous other assessment to guide the company choices.

SAS has a possible to deal with substantial quantity of facts. As a company intelligence tool, SAS empowers the company to effectively deal with monumental quantity of facts and perform assessment on the available data for thousands and thousands of shoppers. Furthermore, SAS with its telecommunication specific answer (TIS – Telecom Intelligence Resolution) assists in building the facts warehouse to hold the expected parameters for further assessment.

As a result, SAS-TIS can be an economical tool for company intelligence pursuits in the telecom sector.

Hyperlink: SAS organization aspects: http://www.sas.com/

Hyperlink: Arindam’s Profile: http://in.linkedin.com/in/arinmukh


About Editorial Staff

The Editorial Staff are the volunteer members behing the Troubled Projects SIG publication. It is formed by top referent experts globally recognized. The Editorial Staff directs and performs quality control and assurance in all the articles published.