P analysis are provided which will help in the

P { margin-bottom: 0.21cm; }A:link { }

Customer
Analytics and Big Data:-
Customers
are considered the heart of every industry. In banking domain, the
most valuable asset is the customer. Customer should be considered as
the focal point of the entire process. Banking Industry is shifting
its focus from products to customers. It is in the race of becoming
customer- centric. With the increasing amount of data generated by
both internal as well as external sources, banks should get an
enterprise view of the customer. This huge amount of amount can help
in the decision making and gaining deeper insights about the
customer. Banks should have a 360 degree view of the customer. With
traditional data collection and analysis methods, even 180 degree
view of the customer was not available. Innovations in technologies
can help the customer to be a touch away from any service provided by
the organization. These touch points are the points of contact when
customer needs a service or support from the organization. The number
of touch points is proliferated which makes it even more difficult to
aggregate and analyse data from heterogenous sources. It is very
hard for the data scientists to mine and analyze the huge mountains
of data accordingly. Different techniques of analysis are provided
which will help in the exploration of useful insights about
customer priorities. These techniques are used in different stages of
customer analytics.
1.
Customer Segmentation
Customer
segmentation or Client Segmentation is the process of the division of
customer sample space into groups or clusters that are related to
each other in a specific manner. The inter and intra relating
attributes may include customer’s bank deposit details, age , gender,
number of dependents, transaction patterns etc. Banks are nowadays
data driven, so it becomes important to collect and gather
appropriate type of data. Data gathering should be taken into
coinsideration as data is arriving from heterogenous sources such as
Internet Banking, Credit card details, ATM transactions, etc. Then
proper method should be developed for analysis. Depending on the
volume, dimension and schema of data, appropriate techniques for
analysis should be used. Clustering is one of the commonly used
techniques for customer segmentation. It is an unsupervised machine
learning technique that tries to find gruopings of similar data
points in the input sample space. This technique outputs a group of
customers who are separated by a varied set of attribute values. The
results of customer segmentation should be communicated properly
among all the applicable domains. The main aim of customer
segmentation is to make organizations understand that every customer
is different from every other customer. It will provide deep
understanding about the demands, priorities and preferences of the
customer. It is obvious that the results of marketing will be
tremenduos if addressed to potential customers. Customer
segmentation defines diiferntiators that partrition clients into
different target groups. Demography is a well known partition
factor. It includes age, gender, religion, race, intellectual level,
family income etc. Other factors include geography, psychography and
behavior. Segmentation of customers help the organizations in cross
selling and upselling of the products. This kind of marketing is
personal to customer adding appreciation and loyalty towards brand
from the customer side. Impersonal and unrelated makrting materials
switches off the interest of the customer.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

2.
Prediction of the customer plans
After
the customer segmentation, next important step is the prediction of
customer actions. Once the understanding of the customer is done, it
will infuse business agility. Segmentation of customers yields
different customer groups. For prediction, all the historical data
related to a particular group should be deeply analysed. More the
historical data available, more accurate will be the prediction. Here
the issue of storing massive data sets come in picture. Earlier as
the cost of storage space was very high, after the specific period of
time, all past records of customers were deleted. This was the major
hurdle in accurate prediction of data.With the evolution of cloud
computing which follows a simple rule of pay as per use helped
organizations to keep their data on clouds in minimum available
prices. The biggest advantage of predicting customer actions is that
it helps to detect and prevent customer churns. It is revealed by the
organizations that the cost of gaining new customers is much more
higher than maintaining the existing customer. Maintining the
existing customer involves a series of steps which include
identifying customers data and its relationships. Data should be
analysed in such a way that correlation between different customers
can be found, if it exists. Different prediction techniques are
available which are used depending on the type of data. For example,
if the analyst is interested in the linear relationship between the
dependent and the response variable, regression is used. If the data
points in the input sample space are multiple, affect of on variable
on all other variables in the input space can be analysed by
techniques like Support Vector Machines and Bayesian networks. If we
are dealing with mixed mode data, decision trees is an appropriate
choice. These techniques will refect the already present outliers in
the data. Depending on the context, either outliers are discarded or
given some special attention. If outliers are reflected in the
network traffic, it reveals the presence of mailicious user in the
network which needs special attention. This leads to the discovery of
frauds in the organization. Frauds are becoming sophiticated so
should be the procedures to deal with them.
Fraud
Detection In Banking with Big Data:-
Analyzing
huge and enormous amounts of data helps in the detection of frauds
in the banking domain. It helps the organisation to fight against the
vulnerabilities by using proper analytics. Banking Industry is most
vulnerable to frauds. If a bank indulges in a fraud at some point,
customers mostly refrain themselves from doing businesses and using
services from the bank in the future. Neither frauds nor fraud
designs and tactics are traditional. Well bred and jaded fraud
programs force organizations to respond in a new and different
manner. Statistics reveals that nearly seventy one percent of
customers switch their banks due to the frauds. Frauds cannot creep
into the organisation suddenly. Banks can analyse the streaming
transactional data in real time.Transactional behavior can help to
detect the unusual behavior of the user which can lead to some
mailicious activity. Banks can analyze petabytes and zetabytes of
historical data which can help in the future prediction accurately.
These analytics can give the patterns of frauds which has happened
earlier. This can help to detect frauds and stop them as soon as
they are about to occur, helping greatly to an organization as it
does not cause any serious damage. The best advantage of detecting
fraud at early satge is its low cost and expenses. Elegance of an
effective analysis is to predict the fraud before it happens or when
it strats happening. After fraud prediction, proper method should be
employed and the roots of fraud should be investigated. These
investigations would turn fraud intelligence into actions. Usage of
more internet banking, credit cards and atm cards open doors to more
production of data. At the same time, it provides a chance of
penetration by the hackers. Undoubtedly, the quantity of the data
produced by the bank is giant and the structure of data poroduced by
banks is complex. Traditionally, future decision making in banks was
improved by digging and mining the already available internal
information about the customer. Internal information and periodic
reports cannot help in leveraging the insights about the frauds and
malicious activitities. Special mechanisms should be employed to
predict, detect and investigate frauds. The mechanisms should be
agile because the situation of organization after fraud pediction is
untenable. This agility in mechanisms is provided by big data
analytics. It monitors the network traffic and system logs
continously. There was a big breach in storing the data
traditionally. User logs and network traffic records were deleted
after a specific amount of time. As a result, organization could
never visualize an overall picture of the transactions of customers .
Big data and cloud computing are two closely inter related terms.
Cloud Computing provides an easy way to save the data on cloud in
mininmum possible rates. This helps in storing all the data without
any need to delete some part of it.

Understanding
customer views

It
is important for the organizations to know how the brand name, brand
image, services and products are percieved by the customer.
Organizations are keen to know whether the customer is satisfied with
the existing services. Organizations are happy to welcome
recommendations from the customer side. Many organizations conduct
surveys online as well as offline to know the customer views. It is
observed that surveys refect what customers say. These statements
could be different from the real actions of customers. Strategies of
the competitators cannot be exactly known. To increase the stability
of the organization in the agile market, it is important to
understand the customers. Customer are the only entities which can
drive the business to unimaginable heights.

x

Hi!
I'm Clifton!

Would you like to get a custom essay? How about receiving a customized one?

Check it out