Big Data for the Next Big Idea in Financial Services (Whitepaper)

Big Data for the Next Big Idea in Financial Services
Understanding Customers, Global Economies and Human Welfare with Analytics

CONCLUSIONS PAPER

Insights from a panel discussion at the
SAS Financial Services Executive Summit

Featuring:

Aditya Bhasin, Consumer Marketing and Online/Mobile Banking
Executive, Bank of America

Robert Kirkpatrick, Director, Global Pulse, United Nations

A. Charles Thomas, Market Research and Analytics Executive,
USAA

Moderator: Jim Davis, Senior Vice President and
Chief Marketing Officer, SAS

SAS Conclusions Paper

Table of Contents
Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
What Exactly Is Big Data?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Beyond Big Data Warehouses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Extracting Business Value from Big Data. . . . . . . . . . . . . . . . . . . . . . . . 3
Big Data on the World Stage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Ten Best Practices for Getting the Most From Big Data. . . . . . . . . . . 4
1. Find Ways to Extract More Value from What You Already Have . . . . 4
2. Bridge Gaps Between Organizational Data Silos. . . . . . . . . . . . . . . . 5
3. Move from Hindsight to Insight and Foresight. . . . . . . . . . . . . . . . . . 5
4. Be Creative Finding New Data Sources. . . . . . . . . . . . . . . . . . . . . . . 5
5. Capitalize on the Insights in Social Media. . . . . . . . . . . . . . . . . . . . . 6
6. Give Customers Value for the Data Value They Offer . . . . . . . . . . . . 7
7. Develop a Customer-Centric View. . . . . . . . . . . . . . . . . . . . . . . . . . . 7
8. Build Bridges Between the Business and IT. . . . . . . . . . . . . . . . . . . 7
9. Establish an Analytics Center of Excellence . . . . . . . . . . . . . . . . . . . 8
10. Prove the Value and Build Incrementally. . . . . . . . . . . . . . . . . . . . . 9
Closing Thoughts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
For More Information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10


Introduction
Fifty-seven million customers, 29 million of them using online banking. Ten-million mobile
banking users, growing at a rate of about 30,000 to 40,000 registered users a week.
Billions of customer interactions each week.

That’s the world of Bank of America. “For us, big data is a way of life,” said Aditya
Bhasin, who manages marketing for the bank’s consumer lines of business, as well as
marketing and operations for online and mobile platforms. “We have somewhere on the
order of 65 petabytes of data about our customers and the interactions that we have
with them.

“We use maybe 1 percent of it for analytics – and even then, there’s too much of the
looking back and not enough of the looking forward.”

How can financial services organizations get more value from the data they already
have? What new data sources should they be capturing, and how can they transform
that data into forward-looking insights?

Those were the questions of the day at the May 2012 SAS Financial Services
Executive Summit. In a one-hour panel discussion at the event, leaders from banking,
insurance and public policy organizations shared their ideas about big data – what their
organizations are doing with it, how to deal with it and what they wish they could do.

• Aditya Bhasin from Bank of America explained how financial services companies
can make better use of the data already being collected to differentiate the
customer experience.
• Robert Kirkpatrick, who leads an information initiative of the UN secretary-
general, discussed how data can be used to better understand changes in global
economies and human well-being.
• A. Charles Thomas, a market research and analytics executive at the insurance
and investment company USAA, described how his company is navigating the
shift to more real-time and predictive analysis.

1


What Exactly Is Big Data?
Big data is a relative term describing a situation where the volume, velocity and variety
of data exceed an organization’s storage or compute capacity for accurate and timely
decision making.

“The definition of big data is all relative,” said Jim Davis of SAS, who moderated the
panel discussion. “It is not so much about the volume of data but about reaching the
point where, for whatever reason, the organization has difficulty getting to it and truly
harnessing it. We wouldn’t be talking about big data if it didn’t represent some sort of
obstacle to the organization.”

It’s an obstacle when data grows beyond the limits and processing capabilities of
relational database management systems. It’s an obstacle when the data is used
primarily to look at the past but not to predict or optimize the future. It’s an obstacle
when it takes hours or days to get the answers to time-sensitive questions. It’s an
obstacle when the window of opportunity – such as the chance to get customer insight
Big data refers to enormity
right to the point of contact – has passed.
in five dimensions:
Volume – From terabytes to
Beyond Big Data Warehouses petabytes and up.
Our thinking about data has fundamentally changed in just a few years, said Davis. Variety – An expanding universe of
“If we were having this conversation five, six or seven years ago, we would be talking data types and sources.
about the importance of consolidating information from all the product silos in our
Complexity – Need to correlate and
financial institutions, building an enterprise data warehouse, and then making sure the
share data across entities.
data is clean and made available for people to apply certain analytics to it so that we
can make fact-based decisions that we hope are predictive in nature. That sounds really Velocity – Accelerated data flow in
good, but we’re past that point today. Now we really need to be looking forward to what all directions.
we can do faster and how we can handle big data without having to bring it all together
Variability – Inconsistent data flows
into an enterprise data warehouse.”
with periodic peaks.
Bhasin agreed: “Historically, the first thing you worried about was, ‘Let’s put everything
into one data mart or one big warehouse, and then we’ll start working on it.’ As more
and more tools and platforms get developed that allow you to use highly distributed,
highly disparate and highly unstructured data sources, we’re able to tie and drive insight
from those sources. That’s where we are excited about big data.”

2


Extracting Business Value from Big Data
The true value of big data lies not just in having it, but in being able to use it for fast,
fact-based decisions that lead to real business value. To that end, the role of data in the
organization has been fundamentally redefined, said Davis. “Traditionally, data was used
in support of business intelligence, OLAP, query reporting, for creating presentations –
those kinds of things. Today, the demands on data are incredibly complex. We’re relying
on data to run our businesses, and we’re using analytics in a predictive way to enable us
to be more proactive in our decision making and less reactive. And the questions we’re
asking are much more complex,” dealing with critical issues of risk, fraud, compliance
and customer insights.

For example, what would it mean to your organization if you could understand The optimistic vision for big
customers not just as segments but as individuals? For example, USAA recently data is that organizations will
expanded its traditional focus on military officers to include members of the military be able to harvest and harness
at any level of honorable service. With more diversity than ever among its 9 million
members, it’s no longer enough to segment by demographics, said Thomas. “Just
every byte of relevant data
imagine, we have a 65-year-old general and we also have a 65-year-old private who and use it to make supremely
served in Vietnam. They have some things in common and other things not in common. informed decisions.
We also have a 19-year-old private. So, traditional ways of looking at our membership
are really having to change.” New data management and analysis techniques enable a Big data technologies now
more granular and real-time view. support the ability to collect
large amounts of data, but more
Big Data on the World Stage
importantly, there’s a shift from
The benefit of big data goes beyond doing what we’ve always done, better and faster.
simply collecting it to being
It is more than doing a better job at the proverbial “right product to the right customer at
the right time.” Big data has a powerful role to play in public trust and industry stability. able to understand it and take
For example, there is speculation that disasters such as the financial meltdown and advantage of its full value – very
mortgage crisis could possibly have been prevented with risk computation on historical quickly.
data at a massive scale.

“We live in a highly connected world,” said Kirkpatrick. “Our economies, our societies,
our markets are connected in ways so complicated we no longer understand them. …
An economic shock in one part of the world can reverberate around the world very, very
quickly. In a matter of months, economic shocks, political unrest and social upheaval can
begin to reverse decades of work that we’ve done to fight hunger, poverty and disease.

“The landscape is so dynamic that you could have emerging vulnerabilities that you
don’t even know about until two years later. That’s a serious problem, because then
we’re finding out today what started happening two or three years ago, because of the
time it takes for statistics to come out. We recognized that we had that challenge, while
ironically, we are swimming in an ocean of digital data.”

3


The ability to rapidly process massive amounts of data is redefining the possible on a
grand scale, said Kirkpatrick. “We see big data as an astonishing opportunity, a really
unprecedented opportunity, to understand what’s happening while it’s still happening,
so we can be more agile developing responses and getting feedback as to whether
those responses are having the desired outcome.”

Ten Best Practices for Getting the Most From Big Data
1. Find Ways to Extract More Value from What You Already Have
“There’s this notion of data intoxication,” said Bhasin. “You can get very drunk on the
fact that there is the other 99 percent of data that’s available, but we’re focusing right
now on making better use of the 1 percent that we do use.”

For example, the bank has transaction records, so it knows where customers shop.
With customers’ permission, could Bank of America use that data to deepen its
relationships with customers? In January 2012, Bank of America started testing a new By finding a new use for
service, BankAmeriDeals, that allows customers to get savings from retailers based on
transactional data the company
their previous spending patterns.
already had, Bank of America
“Most deal services are inherently unintelligent about you,” said Bhasin. “We have the created a new program that
advantage that we know where customers shop, and we can match their shopping
benefits customers, merchants
behaviors to a deal that’s relevant to them, which is populated into their online banking
statements. … The next time you walk into a store, you swipe any Bank of America or and the company. Data that
Merrill Lynch debit or credit product, and the deals automatically apply. You get the cash was formerly just a byproduct
back the same day.”
of the operational environment
The BankAmeriDeals service benefits customers, because everybody loves getting becomes an asset for potential
a deal. It benefits merchants, who can tailor the deals and recipients to influence top-line benefit.
consumer behavior. And it benefits the bank, because it makes the credit or debit card
more valuable – and more frequently used.

“What we’re seeing in the early pilots is that customers who get to know the program
activate their cards more and use their cards more,” said Bhasin. “If there’s some sort
of financial outcomes that occur because of the deals themselves, that’s interesting, but
the real benefit is that our card becomes ‘top of wallet,’ and that’s the holy grail.”

For the United Nations, the holy grail is improving program effectiveness. “We collect
performance metrics across all of the different services and programs that the UN runs
worldwide, and we use this data to understand how well the program worked,” said
Kirkpatrick. “What we’re beginning to realize is that how people participate in programs
changes over time – they use programs differently when their needs change – so there’s
a vast trove of data tracking what people did and how they interacted with our systems
that can tell us more about what was happening in their lives outside of that interaction.
We hadn’t even looked at that before. We suddenly realized this is a dual-purpose data
source, and we hadn’t even begun to touch that.”

4


2. Bridge Gaps Between Organizational Data Silos
When different business functions and flavors of analysis are integrated, the organization
gains essential big-picture perspective, said Thomas. “Under my organization, we
have market researchers and we have what are considered traditional analysts. One
describes the ‘what’ and the other describes the ‘why’ – but together they can say
‘what happens next?’

“We want to examine customer behavior but also to understand the motivation, so we
can either repeat it or stop it, and be able to communicate in a way that resonates.
So bringing those pieces together has been very powerful, because it tells complete
stories. You have full understanding as to why things are happening, so that you can
effect change.”

It is also important to bridge the gaps over time, to connect the dots, said Bhasin. “For
example, a researcher will go out and ask a customer a question, and the customer will
answer the question in the way customers do in focus groups and quantitative surveys.
But as we all know, what they say and what they do are very different things. So you
have to close the loop and see how they actually did behave – and feed that information
back into the next time you ask the question.”

3. Move from Hindsight to Insight and Foresight
“Everybody has analytics today,” said Davis, “but the question is, what sort of analytics
do you have? Do you have analytics that is showing you things that happened in the
past, or do you have analytics that is predictive in nature – such as predictive modeling,
forecasting, operations research techniques for optimization and statistical analysis to
do regression?”

USAA recognized the need to make that shift to a more forward-looking perspective,
said Thomas. “Understanding what happened yesterday really, really well is not going to
be enough for us to operate in real time tomorrow. So starting last year, we’re pivoting,
moving away from what we call hindsight and insight and moving toward foresight.”

Thomas described it as moving from “I know what I don’t know” to “I don’t know what I
don’t know,” to “what if.”

4. Be Creative Finding New Data Sources
Moving up that analytics value chain from rear-view analysis to predictive insight requires
that you use data sources differently or tap into new ones. If you’re just answering
specific questions, the traditional enterprise data warehouse probably suffices, said
Thomas. You can even move into the second realm – finding patterns that were
previously unknown – by delving deeper into the various data sources the organization
already has. But to evolve to the third area – predictive insight – you have to bring in far
more data from external sources.

5


The UN has to be creative to get access to the behavioral data it needs, said
Kirkpatrick. “The best data sources about what’s happening to people are in the hands
of the private sector, and ultimately … that creates fragility for economic and social
shocks. If there were a way for us to create a commons for sharing some of this data in “The best data sources about
ways that protect privacy and competitiveness in the process but still let us make better
what’s happening to people
decisions globally, we could change the world.
are in the hands of the private
“We’re now in discussions with private sector companies across different industries sector, and ultimately … that
about this idea we call data philanthropy. How could you share – not the raw data, but
creates fragility for economic
aggregated data and analysis that powers your business – to reveal patterns, here and
around the world?” and social shocks. If there
were a way for us to create a
With a constellation of data sources, could we identify when people are in trouble,
commons for sharing some of
losing their jobs, getting sick, unable to afford their medicine? Could mobile financial
transaction data show where people are finding employment, where they are in trouble, this data in ways that protect
or when they are starting to default on their mobile phone-based microloan payments? privacy and competitiveness
in the process but still let us
“With these kinds of data sources, we’re not talking about probabilistic models that
something might be happening,” said Kirkpatrick. “This is hard, quantitative evidence of make better decisions globally,
human behavior in time and space. You have to investigate and triangulate to determine we could change the world.”
causality, but being able to actually see collective behavior change using anonymized
data sources is very powerful.” Robert Kirkpatrick
Director, Global Pulse, United Nations

5. Capitalize on the Insights in Social Media
“Social media analytics is a profoundly useful tool for what we’re trying to do, because
you can see intention in a way that we’ve never been able to do,” said Kirkpatrick.
Social media is prevalent even in the world’s developing countries, representing a largely
untapped new data source.

“From the Internet revolution to the mobile revolution to the social network revolution,
populations in places like East Africa have tremendous mobile phone coverage.
Senegal is adding 100,000 Facebook users a month, and there’s a 600 percent
increase in mobile phone coverage in that country. Indonesia is the second-largest user
of Twitter in the world now. There are new opportunities to listen to these signals to help
us fill information gaps in understanding what is happening. If we can figure out ways to
get access to the data, create the right incentive models, do the right analytics and fully
protect privacy in the process, we could have a profound effect.”

Kirkpatrick described a project where social media analysis enabled the UN to predict
unemployment spikes in the US and Ireland four or five months ahead. “We worked
on one project with SAS where we found if you simply analyzed the chatter on
social networks about work in the months before people lose their jobs, they know
something’s wrong and it changes the emotional tone of how they talk about work.”
Imagine the power of being able to drill into particular parts of the country, identify
impending job losses, and then have that advance notice to work with employers and
establish job training programs where they can do the most good.

6


6. Give Customers Value for the Data Value They Offer
There is unquestionable value in being able to analyze and correlate data from a wealth
of new data sources, but at what point does it become intrusive, at least from an
individual perspective? Our panelists agreed on the need to be sensitive to “the creepy
factor” but felt it could be overcome with simple consent. Get customers’ permission to
use the data by offering them some value in return. “In the coming years, we’ll
be looking more at social
“There has been some really interesting research in this area that shows that the
perception of whether or not this is a violation of my privacy is relative to my perception data, more location data, but
of the good, whether it’s a public good or a personal good,” said Kirkpatrick. “Google there are a lot of privacy and
‘flu trends,’ for example. People search for their symptoms on Google; it predicts flu
permission questions that, as
outbreaks by two weeks ahead of CDC, and nobody’s worried about the fact that
it’s aggregated data from search queries. Amazon tells me that people who like this an industry, we all need to be
particular book also like that book. Thank you. That’s a gift; you made my life better careful about before we dig
today. We struggle with this all day, because our mission is to help people, but we’ve got
too deep.”
to make sure that the public good dimension of it, or the personal value, is really clear.”
Aditya Bhasin
Consumer Marketing and Online/Mobile
7. Develop a Customer-Centric View Banking Executive, Bank of America
“We’re very focused on data that can create differentiation in the customer’s
experience,” said Bhasin. “What do we know about either your behavior or your
potential that can cause us to create a meaningfully different environment when you
come and visit us in any one of our channels?” That insight is gained with two primary
approaches:

• Model targeting. “When you come onto BankofAmerica.com, we know who you
are; we know what you have with us. We also know what you don’t have with us
and what you could have with us,” said Bhasin.
• Behavioral targeting. “Based on where you have been on our site, or based
on the product holdings in which you’ve shown interest in a prior interaction,
BankofAmerica.com changes the products that are shown to you in real time.

“These are things we couldn’t do two years ago,” said Bhasin.

8. Build Bridges Between the Business and IT
Our panelists agreed that the need for collaboration is greater than ever. Whatever the
organizational structure, the business people need to understand IT, and IT needs to
understand the business.

“Our best technologists are the ones who sit at our tables, work alongside with us,
and are actively involved,” said Thomas. “It is very important that they’re working with
us right at the beginning, because otherwise there’s too much of, ‘What if we could do
this,’ and then you end up with the business requirements document to fill out, and
then you are by design in an antagonistic environment.”

7


By the same token, the business users need to be technology-aware. “They’re
not going to understand the nuts and bolts of the IT architecture, but they should
understand which systems connect to which systems and how they work, because
then they’re less likely to ask for herculean things that can’t happen.”

The talent model at USAA encourages this business/IT cross-pollination, said Thomas.
“We like to see technologists coming into the business and doing business operating
rules for a while, and going back to technology, and vice versa. We think that’s very
important, because you’ve got to have a core understanding of the business.”

USAA went so far as to create a new career path – basically the equivalent of an
analytics fellow, a role that can go from executive director level to vice president without
managing people. “There will be only a handful of them across the enterprise, but
ultimately we’re trying to create a career path for the best and brightest to not be forced
into managing people.” This represents a big cultural change from the tradition that
equates power with headcount. “Now power is in the ability to make good decisions,”
said Thomas. “We believe that this title, job class and hierarchy are really representative
of what we’re talking about as we move toward being a fact-based enterprise.”

SAS is doing its part by founding a master’s degree program in advanced analytics at
North Carolina State University. “The program has been extremely successful,” said
Davis. “These aren’t people who come from a mathematics or analytics background;
they’re business people who appreciate the need to use what’s available to help solve
a problem. Graduates are getting two and three job offers, and in the last year, the
average salary for a graduate out of this program has been higher than the average
salary for an MBA out of the university, so it shows you how much they are being
valued.”

9. Establish an Analytics Center of Excellence
Since analytics expertise is still in short supply, organizations should take this concept
very seriously, said Davis. “We have help desks in our organizations where if your word
processor is not working or your email is not working or your PC crashed, you have a
single number you can call, and there’s a repository of information and expertise out
there to help. It amazes me that a lot of organizations view data as a strategic asset yet
have no equivalent of the help desk for that environment.”

“At Bank of America we started to have a slightly different problem,” said Bhasin. “We
started to have multiple centers of excellence. Each center of excellence thought it was
individually excellent, and it was, but the problem was that the connections weren’t
being made. The research center of excellence would churn out great research; the
analytics center of excellence was churning out great analytics, and the modeling center
of excellence – and there were two of them, a customer modeling and credit modeling
center of excellence – were each doing great work. Where we were falling short was
that nobody was connecting these pieces.”

8


The company recently consolidated the reporting units, analytics units and modeling
units into a single consumer marketing function, with a chief data officer. “It’s very early
still, but we are starting to see people working together in different ways to come up “If you’re trying to use the ‘I don’t
with common sets of solutions to common sets of problems,” said Bhasin.
know what I don’t know’ or ‘what
if’ mentality, but you’re using
10. Prove the Value and Build Incrementally
methods from the era of ‘I know
IT budgets have certainly not expanded at the same pace as the big data wave. So the
onus is still there to show the ROI for every IT investment, but it can be a challenge to what I don’t know,’ then you’re
show the payback when you’re breaking new ground, said Thomas. “It might not pay really going to be in a position
back anything, but it might pay back tenfold. You never know.” where you never make any
Thomas said USAA is taking an “inch-by-inch” approach. “Rather than go get heaps investments in this space at all.”
and heaps of more data, you have to show the incremental benefit of adding that. If you
A. Charles Thomas
go get data on customer profitability, they say, ‘Okay, I understand, you’re going to drive Market Research and Analytics Executive,
more profitability.’ If you go get customer data on cross-channel behavior, they say, USAA
‘Great.’ You can use it for deflection techniques, and you can put a dollar value on that.
But what’s the value of social media today? We don’t know, but we can test our way
into that.

“So it’s not only the analytical process around how you can wring as much value out of
what you have, but it is also a culture and process change. You have to give the CFO
a reason to believe along the way, show incremental value and partner with IT to drive
that change. But you also have to loosen the criteria for success around IT investments;
otherwise you won’t make any bets in this space at all.”

Closing Thoughts
Big data is not just about helping an organization be more successful – to market more
effectively or improve business operations. It reaches to far more socially significant
issues as well. For example, could we have foreseen the mortgage meltdown, the
financial crises and the recession if only we had gotten our arms around more data and
done more to correlate it? Could we trim millions of dollars in fraud from government
programs and financial markets?

The possibilities are wide open. At SAS, we are optimistic about the potential for
deriving new levels of value from big data with big data analytics. That’s why we
reinvented our architecture and software to satisfy the demands of big data, larger
problems and more complex scenarios, and to take advantage of new technology
advancements.

SAS® High-Performance Analytics was designed to support big data initiatives, with
support for in-memory, in-database and grid computing. SAS OnDemand delivers
any SAS solution on an infrastructure hosted by SAS or private cloud. The SAS High-
Performance Analytics solution on Teradata and EMC Greenplum appliances provides
yet another option for applying high-end analytics to big data.

9


With SAS High-Performance Analytics, the time to decisions can be compressed
from hours and days to minutes and seconds, transforming large volumes of data into L earn More
relevant business value. For example:
For more about SAS® High-
• A large financial services firm with millions of customers reduced the run time for Performance Analytics:
predictive modeling for marketing offers from five hours to less than three minutes. sas.com/high-performance-analytics
• A large investment banking firm cut the time to recalculate the risk exposure of
large portfolios of complex financial instruments from 18 hours to 12 minutes.
To learn about big data:
sas.com/big-data
• A major US bank slashed its loan default calculation time for a mortgage book of
more than 10 million loans from 96 hours to just four.
To read the SAS white paper
Banking on Analytics: How
With that kind of speed, more predictive models can be analyzed, resulting in better
High-Performance Analytics
offer acceptance. Companies can quickly determine exposure and fine-tune their
Tackles Big Data Challenges
responses to market changes. High-risk accounts can be identified early. Customer
sas.com/reg/wp/corp/42594
communications can be framed in the latest knowledge about customer needs or
preferences.
For more about SAS for the
financial services industry:
“The high-performance analytics environment is a real game-changer,” said Davis. “It
sas.com/industry/financial-services
can break down the barriers to big data, and it can help us do things in minutes rather
than hours. Then the obstacle goes away and financial services organizations can focus
on how to capitalize on this power to change the way they do business.”

10

About SAS
SAS is the leader in business analytics software and services, and the largest independent vendor in the business intelligence market.
Through innovative solutions, SAS helps customers at more than 60,000 sites improve performance and deliver value by making better
decisions faster. Since 1976 SAS has been giving customers around the world THE POWER TO KNOW ® For more information on
.
SAS® Business Analytics software and services, visit sas.com.

SAS Institute Inc. World Headquarters +1 919 677 8000
To contact your local SAS office, please visit: sas.com/offices
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA
and other countries. ® indicates USA registration. Other brand and product names are trademarks of their respective companies.
Copyright © 2012, SAS Institute Inc. All rights reserved. 105935_S93315_0812

Big Data for the Next Big Idea in Financial Services (Whitepaper)

More Related Content

What's hot (19)

Viewers also liked (20)

Similar to Big Data for the Next Big Idea in Financial Services (Whitepaper) (20)

More from NAFCU Services Corporation (20)

Recently uploaded (20)

Big Data for the Next Big Idea in Financial Services (Whitepaper)