Scribd
Upload
181 views12 pages

Business Analytics Quest

The document summarizes a case study of a telecommunications company called M>Telco that implemented a business analytics solution to address growing customer disengagement. Key points: 1) M>Telco partnered with IBM to develop a predictive model using customer data to identify at-risk customers and prioritize customer manager assignments. 2) An initial pilot was successful, generating over $7 million in additional revenue. 3) M>Telco then formed a customer intelligence team trained by IBM to expand the analytics program, improving revenue gains and model accuracy. 4) The success prompted M>Telco to fully embrace business analytics across other business areas.

Uploaded by

Jade Mirel Baloloy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
181 views12 pages

Business Analytics Quest

The document summarizes a case study of a telecommunications company called M>Telco that implemented a business analytics solution to address growing customer disengagement. Key points: 1) M>Telco partnered with IBM to develop a predictive model using customer data to identify at-risk customers and prioritize customer manager assignments. 2) An initial pilot was successful, generating over $7 million in additional revenue. 3) M>Telco then formed a customer intelligence team trained by IBM to expand the analytics program, improving revenue gains and model accuracy. 4) The success prompted M>Telco to fully embrace business analytics across other business areas.

Uploaded by

Jade Mirel Baloloy
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 12

Business Analytics Quest

Baloloy, Jade Mirel


Carandang, Jean Paul
Doctolero, Christian Peter
Talingdan, Mark Justin Henry
3ISA

Summary of the Case Study

M> Telco is a subsidiary of a big telecommunications company looking for ways to


make their company grow in a way that is effective both in terms of cost and
benefits. Analyzing their reports exposes the underlying barrier towards their goal,
that is the growing customer disengagement in the midsize group due to
unbalanced allocation of Customer managers to big customers.

In connection to the identification of the problem, the company used business


analytics and partnered with IBM to face the challenge. Conducting seminars by IBM
Educational Service, creating a model to be used, and implementing it are the key
components in solving the problem. This early stage proved to be beneficial as the
company generated more than 7 million in revenue. The solution itself paved the
way for the company's realization on Business Analytics; that is to fully embrace it
and implement on a large scale.

A customer intelligence team was formed and was educated by their chosen
vendor (IBM SPSS) to formulate their own predictive model. The model was
used in a controlled experiment to analyze the benefits of the new approach,
and was successful enough to be implemented for the next succeeding years.

A post implementation analysis was conducted and showed great results. The
room for business growth was expanded and is ready for more challenges to
come.

They later applied the skills and experience from this project to generate
additional returns on other subsequent projects. This major success was
followed up by another seven million increase on their revenue and an
increase in the accuracy of their model by 50% percent. These events
triggered M>TELCO to look into business analytics as an opportunity.

Questions and our Answers


Question 1: What analytics solution was done by the company? How do you
differentiate it with the other types of analytics?

The company devised Predictive Analytics to come up with a solution to their


problem. Predictive analytics is used when you want to forecast the future by
studying recent or historical data. While descriptive analytics reduces huge
amounts of data into a feasible amount of information by answering the
questions of what, where, when and why of a problem. And in prescriptive
analytics, is a suggestion or prescription of a solution based on the forecasted
outcomes of predictive analytics.

Question 2: What are the necessary components in solving the problem?


Early implementation stage
1. They identified the problem, so the scope of the solution would be clear.
2. The study proposes it's solution: Business Analytics. They found out about
Business analytics and explored possible steps in implementing IBM's
predictive analytics.
3. They conducted in-house training followed by a development of statistical
model which analyses the risk of each customer. A gradual transition from
existing system to newer approach is continuing.
4. Based on the model developed by M> Telco team, Customer managers
prioritize customers based on the scores marked with higher priority to be
dealt upon immediately.
5. Post implementation analysis of the implemented system helps the company
realize the need for analytics department, thus customer intelligence team
was formed.

Question 3: Why should analytics be used instead of hiring just more client service
managers?

Both solutions are good but considering the idea of BIG DATA, analytics will be
the better solution in the long run because when it comes to managing
customer satisfaction, huge amounts of customers will also require huge
amounts of client service managers, which is not a good idea considering that
there would be more employees to manage, and more employees to pay
without the assurance of high rates of customer satisfaction.

Predicting the trends will surely increase profitability and retention rather than
passively acting on it, "prevention is better than cure".

Data is already there, waiting to be analyzed. It is practical to analyze it rather


than combating the challenge of retention by force.

Analytics will also be beneficial to other areas, aside from the problem of
customer retention, it can convert data into predictive insights to guide frontline decisions and interactions.

Question 4: Given the CRISP-DM approach that makes use of the following steps in
providing a predictive analytics solution, describe how these phases were done per the
case study given.
a) Business Understanding understanding the nature of the M>Telco is the first step in
this phase. The need of the company to manage their customer disengagement is
one of the barrier they are currently facing. One of their possible solutions is to
increase the number of their client service managers which is considered unwise due
to the fact that it will not give the assurance that it will prevent the customers to
leave their subscription. The other solution that they come up is with the use of
Business Analytics that proves to be effective because it gives revenues to the
company.
b) Data Understanding all the customer data which are too big would be fed to a query
for data mining. Factors are also considered in data gathering, these factors might
include customer satisfaction, package cost, service reliability and quality.

c) Data preparation unstructured data and irrelevant data, are formatted and
selected so that it will be compatible with the model created.
d) Modeling - In this phase, the data set gathered from data preparation were
scrutinized based on the selected factors using the predictive model created by
Customer Intelligence Team. The team analyzed more than 500 variables from
which only the 25 most relevant variables are selected.
e) Evaluation during this phase, the Customer Intelligence Team performed a
controlled experiment to determine the feasibility analytics approach. They have
a control group of customers that are not analyzed using analytics. The result of
the feasibility is successful and can be a repeatable process for addressing high
risk customers.
f) Deployment during 2009, analytics report were made and analyzed by area
for customers outreach. It proves to be a repeatable process ready for full-scale
deployment.

Actions on Disengagement based on Risk Scores


Centralized Retention

Field Retention

Selecting Variables
Actual Variables

Lesser Risk

Excluded Factors

Note: The upper 10% of the scores


are assigned to centralized team,
the next 10% scores are assigned
to field retention team

Question 5: Explain why data understanding and data preparation are important in
this analytics solution.

Huge amounts of varying data are flowing into the company at high speed.
However not all of them are usable to the analyst. Understanding the data is
required to pick out those significant data that is critical for a predicative analytic
solution. Once all of the necessary data is acquired, it must be prepared in a
form of a structured model that will be the basis for the analyst to provide
recommendations to the manager in his decision making.

Financial Analysis
Annual Benefits
Total Benefits Per Period

Indirect

Direct

1000000

2000000

Pre-start

3000000

Year 1

Year 2

4000000

Year 3

5000000

6000000

7000000

Depreciated Assets
Personnel

Consulting

Hadware

Software
0

50,000

100,000
Pre-start

150,000
Year 1

200,000
Year 2

Year 3

250,000

300,000

350,000

You might also like