You are on page 1of 13

3 Top Big Data Use Cases

in Financial Services
How Financial Services Companies are
Gaining Momentum in Big Data Analytics
and Getting Results
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Helping Financial Services Companies

with Big Data

Datameer has been working with some of the most

progressive companies in financial services to rapidly deploy
big data analytics that drive customer acquisition, increase
customer loyalty, predict and prevent fraud and deliver more
targeted advertisement.

The top three credit card companies in the world, top global banks, wealth management
and insurance companies are now understanding their customer behavior by combining
both structured data such as CRM, transaction and point of sales, with unstructured
data such as Web logs, social media and mobile interaction to get new insights that drive
customer acquisition.

PA G E 1
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Get a Competitive Advantage with

Big Data

Financial services companies generate and compile

exabytes of data a year, including structured data such as
customer demographics and transaction history and also
unstructured data such as customer behavior on websites
and social media.

Starting with the financial crisis in 2009, To gain a competitive edge, financial
financial institutions have increased their services companies need to leverage big
focus on customer risk management. data to better comply with regulations,
Around the globe, financial organizations detect and prevent fraud, determine
continue to lay down rules for credit customer behavior, increase sales, develop
risk and liquidity ratio levels, including data-driven products and much more.
regulatory acts such as AML, Basel III
and FATCA that increase the amount of This e-book outlines the three top financial
customer data available for analysis. services use cases of big data as well as
provides a number of real-life case studies.
In addition, financial institutions now have We hope this generates some ideas on
to filter through much more data to identify how you can leverage big data analytics in
fraud. Analyzing traditional customer your organization.
data is not enough as most customer
interactions now occur through the Web,
mobile apps and social media.

PA G E 2
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Understanding Your Customer


As a CMO, digital marketing, or customer loyalty executive

responsible for optimizing customer acquisition and loyalty
campaigns, you need greater visibility into the customer
buying journey.

Why? Because deeper, data-driven

Big Data Analytics at Work
customer insights are critical to tackling Big data analytics is the key to unlocking
challenges like improving customer the insights from your customer behavior
conversion rates, personalizing campaigns data structured and unstructured
Increase Customer to increase revenue, predicting and because you can combine, integrate
avoiding customer churn, and lowering and analyze all of your data at once to
customer acquisition costs. generate the insights needed to drive
customer acquisition and loyalty. For
But consumers today interact with
example, you can use insights about the
companies through lots of interaction
customer acquisition journey to design
Increase Revenue
points mobile, social media, stores,
per Customer campaigns that improve conversion rates.
e-commerce sites, and more which
Or you can identify points of failure along
dramatically increases the complexity
the customer acquisition path or the
and variety of data types you have to
behavior of customers at risk of churn to
aggregate and analyze. Think Web logs,
proactively intervene and prevent losses.
transaction and mobile data. Advertising
Decrease And you can better understand high-
Customer social media and marketing automation
Acquisition Cost value customer behavior beyond profile
data. Product usage and contact center
segmentation (for example, what other
interactions. CRM and mainframe data.
companies they shop from, so you can
And even publicly available demographic
make your advertisements even more
data. When all of this data is aggregated
and analyzed together, it can yield insights
Reduce you never had before for example, who
Customer Churn
are your high-value customers, what
motivates them to buy more, how they
behave, and how and when to best reach
them. Armed with these insights, you can
improve customer acquisition and drive
Enhancement customer loyalty.

PA G E 3
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Fraud and Compliance

If you are responsible for security, fraud prevention, or

compliance, then data is your best friend if you can use it
to identify and address issues before they become problems.
The fact is, security landscapes and compliance requirements
are constantly evolving, as are the methods that the bad
guys are using to defraud your business and customers.

Data-driven insights can help you uncover Big Data Analytics at Work
whats hidden and suspicious and
Big data analytics enables you to
in time to mitigate risks. For example,
combine, integrate and analyze all of your
analyzing data can help you reduce the
data at once regardless of source, type,
Cyber Attack operational costs of fraud investigation,
Prevention size, or format to generate the insights
anticipate and prevent fraud, streamline
and metrics needed to address fraud
regulatory reporting and compliance (for
and compliance-related challenges. For
instance, for HIPPA), identify and stop
example, you can perform time series
rogue traders, and protect your brand. But
analysis, data profiling and accuracy
this requires aggregating and analyzing
calculations, data standardization, root
data from a myriad of sources and types
Regulatory cause analysis, breach detection, and
Compliance and analyzing it all at once no small
fraud scoring. You can also run identity
task. Think financial transaction data,
verifications, risk profiles, and data
geo-location data from mobile devices,
visualizations and perform master data
merchant data, and authorization and
submission data. Throw in data from
lots of social media channels and your
Behavior mainframe data, and you have a significant
challenge on your hands.

However, with the right tools, this melting

pot of data can yield insights and answers
you never had before insights you can
Credit Card use to dramatically improve security, fraud
Fraud Detection
prevention, and compliance.

PA G E 4
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

EDW Optimization

Enterprise data warehouses (EDW) are critical for enabling

operational reports for businesses but as the size and
complexity of the data to be analyzed increases, youll
eventually hit the limits of traditional data warehouses.

Youll know it when your processing produce analytics from unstructured data
times take too long to meet business because they dont support it. This forces
needs, your costs get out of control, or technicians to manually give this data
you struggle to process and analyze new structure before analyzing it.
Expensive data types. For both IT executives and key
stakeholders responsible for analytics, Big Data Analytics at Work
business intelligence and enterprise The good news is, big data analytics
data, this is a serious problem. Todays solutions that run on Hadoop can solve
business decision makers simply cant these challenges. Big data analytics
afford delays in insights anymore. solutions running on Hadoop make it easy
Expensive The solution is to offload the most to overcome these challenges because
Data Preparation
to Lower Cost challenging data management and they allow you to cost effectively scale
analytics activities to new technologies to any volume of data and store and
and management approaches designed analyze any and all data types together
to handle them. For example, do you both structured and unstructured. You
need to cut the costs of data preparation can also use them to extract structured
and cleansing? Reduce time to insight data from your EDW into Hadoop for
Data Discovery
by offloading the most time-consuming cheaper storage and then send back
analytical tasks? Support a variety of new into EDW for analytics. All data can be
data types, especially unstructured data? analyzed as is, eliminating costly data
Or better manage rapidly growing log, preparation activities. At the same time,
sensor and other unstructured data? big data analytics is so powerful because
Deal with Various
Data Types it enables you to combine, integrate and
Traditional EDWs were never designed to quickly analyze all of your data at once
solve these types of challenges. First, they regardless of source, type, size, or format
make it prohibitively expensive to manage to generate the insights your business
the ever-increasing volumes of transaction needs. In addition, you can parse, clean,
and interaction, mobile data, website profile, match, enrich, aggregate, and
click stream data, ad click through, log normalize data, as well as manage ETL
data, sensor data, and unstructured workloads and generate master data.
machine data. At second, theyre slow to

PA G E 5
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Datameer Delivers Faster Insights from

Big Data

Datameer provides a one-stop-solution for getting all of

your Web, advertising, mobile, social media, transaction,
marketing automation, and CRM data into Hadoop; enriching
it with third-party data; analyzing your data; and visualizing
results using wizard-led data integration, point-and-click
analytics functions, and drag-and-drop visualizations.

Our broad set of 60+ data connectors and analytic functions make it easy to:

Rapidly combine and enrich existing data sets with third-party, customer, and other data 

Perform ad-hoc analytics to test what-if scenarios

Better understand your customer interactions and identify the most effective campaigns

Identify the behaviors of customers at risk of churn

As a result, you can answer questions like:

Whats really happening across the customer journey?

Which campaign combinations accelerate revenue?

Which offers drive customer loyalty?

What credit card behavior signals potential fraud?

How can we assess customer risk before extending credit? 

How can create more targeted campaigns?

The following case studies outline how some of the most progressive financial services
companies are using big data for a competitive advantage.

PA G E 6
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Predict and Prevent Customer Churn

The Need for Insight into Customer Analyzing Customer Activity

Behavior Leading Up to a Withdrawal
As part of a strategic initiative to reduce Using Datameer, the company first
C O M PA N Y P R O F I L E : the amount of funds being transferred integrated data of customer activities
A leading financial services to competitors, this financial services that might indicate that the client was
company offering retirement company wanted to understand the client considering moving their funds elsewhere.
planning, IRAs, mutual funds,
behaviors that might signal movement of Examples of such activities included: the
life insurance, and more.
funds. With a growing number of clients client had recently called in for information
retiring, they realized that it was important with an outside financial consultant on
to provide better wealth management the line; a change in address, workplace,
services to retain them. They also wanted or power of attorney; or the client had
to identify and track specific behaviors recently been browsing on the company
so they could focus on the people who site for forms. They pulled multiple data
might consider moving their money. This sources together to build out activity
is important because historically, 50% of paths for each client. For example, they
the time, they can retain a customer if they tracked clients specific activities and
know the customer is considering moving whether these activities ultimately led to
their pension. They leveraged Datameer a transfer or withdrawal. By correlating
to gain insight on what customer behavior this data, they were able to determine the
signals that a customer is likely to transfer statistical relevance of each activity, or
their investment. combination of activities that predicted
customer churn.

By identifying patterns of customers at risk of
churn and proactively reaching out to offer their
wealth management products, the company has
reduced customer churn by 50%.

PA G E 7
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Analyzing Effectiveness of a New Sales

Tool with Behavior Analytics

Measuring the Impact of a New Analyzing Behavior

iPad Application Using Datameer, the team integrated
C O M PA N Y P R O F I L E :
In the United States, the asset structured and unstructured data from
A major provider of worldwide
investment management management industry oversees the multiple internal and external data
services. allocation of approximately $53 trillion in sources. This included data from the CRM
financial assets3. This company wanted system logs, iPad application JSON logs,
to increase sales to get a bigger piece web activity, and the sales database. The
of this large market. They also needed CRM system logs recorded the actual
a way to measure the effectiveness of date and time of the meeting, along with
different sales tools to accomplish this details of what products were discussed.
goal. The product management team The JSON logs showed whether the
created an iPad application that explained application was in use at the specific time
the various fund options and each funds of a meeting between a sales rep and
historical performance. The sales reps a financial adviser. The JSON logs also
used the iPad application during meetings showed the actual fund details that were
with financial advisers (wholesalers) to viewed on the iPad application during the
help describe the fund details. Product meeting. Weblogs showed if the sales rep
management wanted to measure the value browsed through the company website
of the new iPad application by analyzing during the meeting. The sales database
the impact of its use on new trades (sales). showed whether a specific meeting
resulted in the client performing a trade
within two weeks. The sales database also
showed if the trade was for the same funds
that were browsed during the meeting.

These data sets were integrated and

analyzed to show the percentage of
Result the overall meetings where the iPad
application was used, and which meetings
The analysis through Datameer showed that where the iPad application was used
using the iPad application resulted in a significant resulted in trades facilitated by specific
increase in sales within two weeks of meetings. It financial advisers within two weeks.
also showed many of the most successful sales After the initial analysis, the company
reps used the application. conducted a further analysis to determine
whether the most successful sales
reps used the iPad application more
consistently than others.

PA G E 8
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Analyze 7x More Customer Attributes

96% Faster
Ultimately, the bank wanted to accelerate
Process Billions of Data Points
customer spend pattern signature
Banks need to be able to quickly
detection, a computationally expensive
identify potentially fraudulent credit card
analytic process. The goal was to analyze
C O M PA N Y P R O F I L E :
transactions. This requires the ability
historical transactions to create signature
A major worldwide bank,
to swiftly prepare data regarding these
serving individual consumers, vectors for different entities (customers,
businesses of all sizes, transactions and historical spending
clients, industries, branches, economies,
governments, and other patterns. This leading bank has billions
etc.) that capture the norms observed in
of data points to process. Their previous
prior behavior. In addition, they wanted to
analytic environment had limits in
identify changes in behavior by classifying
computational power and in the ability
current transactions and comparing them
to handle large volumes of data. The
to the relevant signature vector.
company wanted to increase the speed,
agility and capacity to build, refine, and
Analyzing Millions of Credit Card
calibrate analytic models with three
specific goals: eliminate sampling bias by
The company built a query as a Datameer
building models with more data; test and
workbook for millions of credit card
iterate models faster and increase overall
accounts, encompassing almost three
performance as compared to existing
billion transactions; 5 billion customer
models while improving modeler efficiency
attributes; 50 spending categories; and 55
and productivity.
attributes per category.

Using Datameer, the company now analyzes
5 billion customer attributes in 51 minutes. In
comparison, the company using their existing EDW
solution, could only analyze 700 million customer
attributes in 36 hours.

The Datameer solution provided a significant

improvement in performance, reducing processing
time by 96 percent while processing 7 times the
number of customer attributes.

PA G E 9
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Reducing Customer Acquisition Costs

with Targeted Promotions

The High Cost of Acquiring New Using Insight for More Targeted
Customers Promotions
A McKinsey Global Institute Report This company used Datameer to correlate
shows that marketing and sales consume customer purchase history, customer
C O M PA N Y P R O F I L E : about 15 percent of costs for bank and profile data, and customer behavior on
A leading multinational financial insurance companies. Institutions spend public social media sites that indicated
services provider based in the
hundreds of dollars to acquire each new areas of interest (for example liking
United States
customer. Credit card issuers spend something on Facebook). Using the insight
billions of dollars annually on direct mail gained, the company could offer special
and mass-market advertising to get new promotions to an individual based on their
accounts. 4 interests. For example, if a person made
a number of transactions at Whole Foods
This company realized that it needed
and liked the Food Network, the company
to find a way to target customers and
could send a credit card with a special
promote to them more effectively to
promotion for Whole Foods or the Food
reduce customer acquisition costs.

Using the insight provided through the Datameer
analysis, the company created special targeted
promotions, reducing the cost of new customer
acquisitions by 30%.

The social economy: Unlocking value and productivity through social technologies, July 2012
PA G E 10
Datameer 3 T O P B I G D ATA U S E C A S E S I N F I N A N C I A L S E R V I C E S EBOOK

Ensuring Compliance with Basel III

through Analysis of Trillions of Records

The Importance of Data Accuracy changing adversely. In addition, they

and Quality want to make sure that the risk profiles
C O M PA N Y P R O F I L E :
Basel III has introduced a number of being reported to investors and regulatory
A leading retail bank agencies are prudent and in compliance
challenges to banks. An Accenture report
states, The technical challenges of Basel with regulatory requirements.
III implementation includes data availability,
Prior to Datameer, the bank used a SAS
data completeness, data quality and data
application to analyze data quality. They
consistency to calculate the new ratios.5
also were using Teradata and a mainframe
This bank integrated data from multiple solution. The process was time consuming
sources for over 300 attributes. The data and complex.
sources included sub-prime loans data,
acquired wealth management data, and Quick Analysis of Trillions of
data from legacy retail banking systems. Records
To comply with Basel III, the bank needs Using Datameer, the Data Quality Initiative
to ensure quality of the data, through a team analyzes trillions of records that
scorecarding process. The scorecarding result in approximately 1 terabyte of
process includes subjecting the data to reports per month. The team puts the
hundreds of data quality checks. The results of the analysis in a data quality
results of those checks are trended over dashboard that the company uses to
time to ensure that the tolerances for ensure accuracy of regulatory compliance
data corruption and data domains arent reporting. In addition, using Datameer,
they are able to reduce the time it takes to
determine the impact of the data on risk
metric calculations and accelerate time to
market of their risk analyses.

With Datameer, the bank has dramatically reduced
the time to analyze hundreds of attributes and
terabytes of data.

5 Consequences.pdf
PA G E 11

2016 Datameer, Inc. All rights reserved. Datameer is a trademark of Datameer, Inc. Hadoop and
the Hadoop elephant logo are trademarks of the Apache Software Foundation. Other names may be
trademarks of their respective owners.

1550 Bryant Street, Suite 490 San Francisco, CA 94103 USA Tel: +1 415 817 9558 Fax: +1 415 814 1243

9 East 19th Street, 5th floor New York, NY 10003 USA Tel: +1 646 586 5526

Datameer GmbH Magdeburger Strae 23 06112 Halle (Saale), Germany Tel: +49 345 2795030

Datameer Singapore Pte Ltd 03-20 Galaxis 1, Fusionopolis Place Singapore 138522, Singapore Tel: +65 6809 1157

12/F International Commerce Centre, 1 Austin Road, West Kowloon, Hong Kong Tel: +852 2824 8646