Description
A smarter planet creates new possibilities, as well as new complexities and risks. Thanks to recent technological advancements, including mobile devices, cloud computing and social media, we have created an interconnected, instrumented and intelligent world in which there is virtually infinite access to information.
IBM Software Predictive Analytics Software
Business Analytics
Predictive Threat
and Fraud Analytics:
Meeting the Challenges
of a Smarter Planet
2 Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
Contents For example:
2 Introduction
What is business analytics
!educe exposure and minimi"e impact
!2 Five steps to a proactive threat and ris# strate$y
! %rive insi$hts with IB& Business Analytics solutions
!" 'onclusion
!" About IB& Business Analytics
Introduction
A smarter planet creates new possibilities( as well as new
complexities and ris#s) *han#s to recent technolo$ical
advancements( includin$ mobile devices( cloud computin$ and
social media( we have created an interconnected( instrumented
+
+
+
+
+
Accordin$ to a recent study by the Association of 'ertified
Fraud ,xaminers( the typical or$ani"ation loses - percent of
its revenue to fraud every year) .lobally( it/s estimated that
annual fraud losses exceed 01)- trillion)2
3early half of or$ani"ations that are victimi"ed by fraud are
unable to recover their losses)2
,stimates from 3ational 4eath 'are Anti5Fraud Association put
health care frauds costs in the 6S at 078 billion annually)
9ther estimates ran$e as hi$h 021: billion)1
*he total cost of insurance fraud ;non5health insurance< is
estimated at more than 0=: billion annually( costin$ the
avera$e 6S family up to 0>:: in increased premiums(
accordin$ to the FBI)=
.lobally( the total cost of credit card fraud is estimated to be
0-)- billion) In the 6S alone( ten percent of citi"ens have
been victims of credit card fraud)-
and intelli$ent world in which there is
virtually infinite access
to information) While this has opened the door to new and
excitin$ possibilities( it also provides an openin$ for new
threats and vulnerabilities) From external factors such as
national security and the economy ? to internal factors such as
insider fraud( credit ris# and information mana$ement ?
or$ani"ations face a multitude of threats every day) *hese
threats are increasin$ in number and severity( and can cost
or$ani"ations millions( even billions( in losses)
It is therefore not surprisin$ that threat and fraud miti$ation
has become a top priority for business) In addition to the
monetary losses mentioned above( there can also be a
si$nificant cost to the reputation and value of an or$ani"ation
that is unprepared for threats and fraud) Accordin$ to a recent
IB& $lobal study of senior executives( >- percent say that data
theft and cyber attac#s impact customer satisfaction and brand
reputation) And 72 percent say they are they are the $reatest
threats to their company/s reputation)7
IBM Software 1
*he best defense a$ainst threat and fraud is a systematic
approach to reducin$ exposure and minimi"in$ ne$ative
impact) But where do you be$in In an ever5chan$in$ world(
you must be dili$ent about mana$in$ the threats you are aware of(
as well as those yet to be identified) For example( you need
to be able to anticipate how an insider mi$ht infiltrate your
secure I* systems( loo# for si$ns of potential terrorist
activities( predict the li#ely impact of future economic events
or identify new patterns of fraud ? all with a hi$h de$ree of
accuracy)
In this paper we will discuss how to build a proactive threat
mana$ement strate$y based on business analytics technolo$y(
and then deploy that information to the individuals responsible
for impactin$ unwanted behavior) @ou/ll see examples of how
or$ani"ations worldwide are applyin$ business analytics
solutions to minimi"e the ne$ative impact of ris# and maximi"e
positive results) @ou will also learn the practical steps you can
ta#e to combat threat and fraud in your or$ani"ation)
What is business analytics
*he term Abusiness analyticsA refers to a comprehensive(
data5driven approach that combines the science of predictive
analytics with advanced business intelli$ence capabilities)
Predictive analytics uses advanced analytical al$orithms to
process historical data and create models that can ma#e
predictions about future outcomes) Business intelli$ence
capabilities then deliver these predictive insi$hts to #ey
personnel and departments across the or$ani"ation to help
them achieve their $oals and obBectives)
,ach of these technolo$ies is powerful in and of themselves)
When combined as business analytics( they provide superior
capabilities for improvin$ business performance across a wide
ran$e of functional areas) Accordin$ to analyst firm I%'(
or$ani"ations that utili"e predictive analytics in addition to
business intelli$ence achieve an avera$e return on investment
of 2-: percent)>
&odern business analytics software is easy to use and can be
operated by users with different s#ill levels( from front5line
business users to experienced analysts) Businesses across many
industries use business analytics to understand their customers(
increase profitability and improve operational efficiency and
effectiveness) When applied to threat and fraud prevention(
this technolo$y provides similar benefits by providin$ the
ability to connect data to effective action by drawin$ reliable
conclusions about current conditions and future events)
!educe exposure and minimi"e impact
&ountin$ re$ulatory demands( the $rowth of online
transactions and communication( the increased use of mobile
devices and the constant shadow of an uncertain economy
underscore the importance of proactively mana$in$ threats in
all their forms ? whether related to business( data or events)
*hat means not only understandin$ what has happened in the
past( but bein$ able to loo# forward to anticipate what mi$ht
happen in the future( and how your business will respond to
these situations) 9r$ani"ations that do this well can
si$nificantly maximi"e efficiency and reduce the impact of
threats( as well as differentiate their business model from those
who have not adopted a culture of analytics)
= Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
Business analytics is the #ey to reducin$ exposure and
minimi"in$ the ne$ative impact of the thousands of threats
your or$ani"ation faces every day) It provides you with access
to fact5driven predictive insi$hts in real5time( driven by your
or$ani"ation/s specific needs)
*oday/s dynamic environment reCuires a systematic approach to
mana$in$ threats) By usin$ business analytics throu$hout your
decision5ma#in$ lifecycle( you can continuously refine
the decisions you ma#e and the strate$ies you use as an
or$ani"ation to ta#e control and ta#e actions that are based on
insi$ht) As illustrated in Fi$ure 2( when you apply business
analytics( you become empowered to:
As you can see( this cycle is continuous: as you acCuire new
data( you analy"e it and learn from it) *he insi$hts you $ain
then help you improve policies and parameters so that your
or$ani"ation is constantly shielded from ne$ative ris# at
each sta$e)
%efine
Allow
+
+
+
%efine adaptive( forward5loo#in$ policy definitions and
parameters based on past actions( new challen$es and
internal or external chan$es( which can include anythin$
from the weather to evolvin$ re$ulatory reCuirements)
&onitor your environment continuously usin$ a broad ran$e
of data from multiple sourcesD addin$ new data as needed
and learnin$ from your data to develop insi$hts so that you
can identify tri$$ers and alerts)
%etect suspicious behavior such as threats( information
breaches( crime and fraud( usin$ business analytics to
Prevent
%etect
&onitor
identify anomalies and determine the li#elihood that an Figure 1: *oday/s dynamic environment reCuires a systematic approach to
mana$in$ threat and ris#)
action poses a potential ris#)
+ Prevent or allow a potential threat based on its anticipated
impact on your or$ani"ation( and either eliminate or mana$e
the ris# to ensure a controlled( output5driven approach that reduces
exposure or loss and maximi"es the positive impact
of any action ta#en) For example( a ban# can analy"e a credit
card applicant/s information( includin$ credit score( marital status(
lon$evity at current employer and current salary( to
determine how these factors into the relative ris# of default) *he ban#
may determine that while the li#elihood of default isn/t hi$h enou$h
to prevent it from securin$ the credit card(
they cannot afford to issue the card at a low interest rate( thereby
miti$atin$ some of the ris# it is assumin$)
IBM Software -
&asterin$ business analytics is a Bourney that starts with
usin$ the information you have to find insi$hts( and continues
throu$h determinin$ the next best action for a particular
#ith a $usiness analytics%$ased ris& management strategy
in 'lace you $ecome a$le to:
individual( scenario or decision ma#er) By applyin$ business
analytics techniCues or$ani"ations are able to identify #ey
threats( accommodate re$ulations( or refine and monitor
policies based on what they #now about past events as well as
data obtained in real5time)
*hat can mean( for example:
+
+
+
+
+
Access threat information across the organization in real time
to determine the best time to accept the risk or stop it
altogether.
Make decisions faster and spend less time "putting out fires."
Gain more confidence and trust in decisions that are made.
Reduce costs and make better use of resources.
Build a stronger organization that is resilient to change and
read to e!ploit new opportunities.
+
+
%eterminin$ that an auto insurance or medical claim is
fraudulent as it is processed)
Sendin$ police into areas most li#ely to have hi$h crime
activity on a particular day based on #ey predictors such as
the weather( previous days/ behavior( day of the wee# and
+
+
Support business inno"ation while reducing or maintaining
an acceptable le"el of operational risk. A"oid negati"e
publicit.
+
+
other factors)
Identifyin$ the person in your enterprise who suddenly
seems to be en$a$ed in irre$ular networ# and data activity)
Providin$ current( customi"ed threat intelli$ence reportin$
to specific employees across the or$ani"ation so they can
anticipate and address ris#s within their areas of
responsibility)
Identify organi(ational threats with
$usiness analytics
Patterns that point to threats or suspicious behavior are
often hidden in hu$e amounts of data) *his is why a proactive
approach to threat mana$ement starts with analy"in$ the
masses of data your company collects and stores such as
transactional data( email( networ# reports( survey data(
constituent information( tax records and call center notes) It/s
4avin$ the answers ahead of time $ives you the ability to
control what action you ta#e and when so that you can plan(
forecast( implement and prevent) Without analytics( you can
react only when a threat is identified( which puts your
or$ani"ation into Afi$ht or fli$htA mode( instead of enablin$
you to respond in a controlled and formulated way that will
ensure the best outcome) *his can be very costly to your
or$ani"ation ? costly in terms of lost revenue or opportunityD
reduced credibility in the eyes of employees( partners and
investorsD ne$ative environmental impactD or even lives lost)
eCually important to monitor and analy"e the massive amount
of information flowin$ throu$h social media sources such as
networ#in$ sites( blo$s and comment sections of websites)
7 Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
Business analytics solutions feature sophisticated techniCues
that enable your or$ani"ation to analy"e both the AstructuredA
data found in tables and databases as well as AunstructuredA
text( includin$ content found in email chains( chat forums(
social networ#s or survey data) In contrast to rules5based
analysis and detection methods alone( business analytics can
identify relatively unusual behaviors( even those with subtle
differences that other methods often miss) And because new
schemes continually emer$e and chan$e over time( this
process also helps identify new threats)
*his allows or$ani"ations to combine a diverse ran$e of data
types( includin$ data that is specific to your industry in terms
of benchmar#s or external data) Business analytics $ives you
the ability to combine a wide variety of data dimensions( types
and sources on an on$oin$ basis) *his ma#es it possible to
Cuic#ly and reliably detect inadvertent si$natures from hac#ers(
criminals or terrorists $eneratin$ new cyber chatter or tryin$
new tactics to $ain access to sensitive information( or to
determine the ris# of $rantin$ an individual a credit line)
%ifferent techniCues can be used to help you decide which
next best action you should ta#e)
9nce you have an understandin$ of your data( includin$ what
constitutes normal and unusual behavior( you can develop #ey
predictors( or indicators of potential threats( that can be used
alon$side your #ey performance indicators to hi$hli$ht areas
of concern for your business or or$ani"ation) @ou can then
be$in to base strate$ic and operational decisions on predictive
intelli$ence that not only shows future outcomes and behavior(
but also details the factors that influence those events) *his
increases your ability to proactively stop threats before
they occur)
Business analytics also enables or$ani"ations deliver the models
and business rules that support consistent( automated decisions
at the Apoint of impact)A *his could mean deployin$ analytic
results directly to #ey people throu$h business intelli$ence
reports( dashboards or other meanin$ful visuali"ations such as
$raphs or maps) ,mployees receive customi"ed reports that
address their particular area of concern( and are able to drill
down into the data to find root causes and $ain deeper insi$hts
for predictin$ and preventin$ the ne$ative conseCuences of
threats)
*hese insi$hts can be also embedded within operational
systems such as websites and call centers to drive smarter
decisions and actions that are ali$ned an or$ani"ation/s overall
strate$y and $oals) *his capability provides a critical layer of
optimi"ed decisions where a company/s staff or automated
systems interact with customers( prospects( suppliers or
partners) Information about results is then used to refine
predictive models and future recommended actions)
Eet/s loo# at some areas in which business analytics5based ris#
and threat control strate$ies are particularly effective)
)liminate insider threats
For most or$ani"ations( data has become an invaluable
asset ? the lifeblood of their operations) Access to this data is
available to an expandin$ user base( includin$ employees(
business partners( suppliers and customers) I* infrastructures
are more extensive( more complex( more distributed ? and
more accessible) *his interconnectedness affords many benefits
for businesses( $overnment a$encies and consumers ali#e ? but it
also introduces a $reat deal of potential ris#) 'ompanies and
$overnment a$encies must answer to strin$ent re$ulatory
reCuirements and protect intellectual capital from competitors
or subversive political entities) And consumers are typically
most concerned with the potential for identity theft and other
privacy violations)
IBM Software >
*he Wi#iEea#s scandal involvin$ the 6S State %epartment is
an ideal example of one of the primary challen$es to data
security: the authori"ed insider threat) In this instance( massive
amounts of secret documents ? includin$ thousands of
embassy cables and hundreds of thousands of documents
relatin$ to the war in Af$hanistan and IraC ? were copied by
authori"ed users to '%s and %F%s and sent to Wi#iEea#s)
Intrusion analysts at a national $overnment a$ency are often
flooded with several million events per day ? many of which
are mappin$ reCuests or information $atherin$( rather than
actual threats) *his lar$e number of events ma#es detectin$
real attac#s or potential problems difficult) 6sin$ IB&
Business Analytics software( analysts can focus on the alarms
most li#ely to be cause for concern( in one instance reducin$
the number of events reCuirin$ manual review by G> percent in
a sin$le 1:5day period)
Authori"ed insider threats are not uniCue to the $overnment or the
military) Firtually any or$ani"ation that possesses sensitive
business information such as earnin$s releases( mer$er and
acCuisition plans( mar#etin$ plans( research and development
data( personally identifiable information( or sensitive internal
emails are at ris#)
9ne of the $reatest areas of concern re$ardin$ the breach of
sensitive data is within the healthcare industry) *here have
been a number of hi$h profile cases recently in which personal
health information was stolen due to insufficient safe$uards)
For example( the 6tah %epartment of 4ealth reports that
social security numbers( addresses( dia$noses( dates of birth
and other confidential information for >8
::: patients was
compromised durin$ a hac#er attac# in 2:22) *his is not an
isolated incident) .overnment fi$ures show that hundreds of
thousands of patient records have been stolen from a variety
of healthcare or$ani"ations)8
Both the Wi#iEea#s and healthcare breaches demonstrate the
value of implementin$ business analytics technolo$y for the
prevention of data theft)
*educe credit card ris&
In today/s turbulent financial mar#et( credit card issuers need
to #now as much about their customers as possible) *hey need to
understand the factors and behaviors that indicate whether a
potential customer is a $ood credit ris#( as well as patterns or
trends that point to the probability a customer or applicant will
miss payments or default) In addition( they need to devise
more sophisticated strate$ies to #eep the $ood customers and
limit the ris#s associated with unprofitable ones) IB& Business
Analytics software allows or$ani"ations to build credit models
that are based upon a more refined se$mentation strate$y that
ta#es into account customers/ uniCue and individual data(
ensurin$ that credit policies are ali$ned with customers/ level of
ris# to ensure the most profitable outcome)
A private and independent ban#in$ institution in Swit"erland
improved the performance of its credit ris# mana$ement
system by implementin$ a new credit ris# scorin$ system based
on IB& Business Analytics software) *his solution enabled the
ban# to improve the mana$ement of credit ris# and the overall
collection procedure( and has reduced unpaid debts by 2- percent)
8 Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
+etect and 'revent fraud
Fraud is an expensive problem that costs industries and
countries billions of dollars) *he ran$e of fraud( and the
resourcefulness of fraudsters( poses a dauntin$ problem for
those char$ed with its detection and prevention) @et these
criminal activities have one thin$ in common: each of them
leaves a trail of behavioral and transactional data ? insurance
claims( mort$a$e or benefit applications( healthcare
submissions( tax returns and so on ? which are filed alon$side
information from le$itimate interactions) 'onventional
analytics can be used to detect fraud after it has happened)
IB& Business Analytics software can forecast what is li#ely to
happen in the future( as well as detect events as they happen)
'onsistent fraud intelli$ence is provided to the people on the
front lines of the business to catch fraud in the act( or even
prevent it before it ta#es place)
Minimi(e inventory loss
Business analytics helps identify products( store conditions(
personnel or customers who may be lin#ed to stoc# shrin#a$e)
Shrin#a$e solutions from IB& provide retailers with insi$hts
into product and store performance around stoc# control and
fraudulent activities( and can help increase the profitability of
store operations)
Assess networ& outages
Business analytics can help telecommunications companies or
lar$e retailers improve networ# asset mana$ement outcomes
by predictin$ which networ# will fail next( and the impact
this will have on operations and customer experience) IB&
Business Analytics solutions ensure that or$ani"ations can ta#e
Cuic# action to repair at5ris#( hi$h5traffic networ#s( enablin$
them to avoid the ris# of downtime( as well as customer
dissatisfaction and lost revenue)
+ *he anti5fraud a$ency of a southern ,uropean country
implemented IB& Business Analytics technolo$ies to Prevent energy fraud
,ner$y fraud can be difficult to detect and cause maBor
si$nificantly improve the speed( effectiveness and ease of
financial losses for ener$y a$encies and hi$her prices for
fraud detection than#s to accurate( automated identification
consumers) IB& Business Analytics solutions help prevent
of hi$h5ris# taxpayers)
ener$y fraud and theft by detectin$ suspicious activity based on
+ Eeadin$ healthcare or$ani"ations around the world use
distribution( consumption( pricin$ and smart meters and then
IB& Business Analytics solutions to minimi"e the impact of
determinin$ appropriate intervention strate$ies) *hey also
fraudulent claims by ensurin$ early detection of li#ely
help a$encies better allocate their investi$ative resources(
instances before payment occurs)
predict hi$h5ris# areas for fraud and theft( anticipate
+ 'laims and warranty departments can reduce costs and
recurrence ris#s and increase customer satisfaction
improve supplier relationships by redirectin$ fraudulent
decreasin$ the costs associated with ener$y theft)
warranty claims and presentin$ this information via
reportin$ methods inside of claims systems)
IBM Software G
Protect national $orders
Protection a$ainst threats often be$ins at border crossin$s(
airports and in harbors) IB& Business Analytics solutions help
a$encies identify which containers enterin$ a port could
contain unwantedHdan$erous materials( which passen$ers on an
airline should be investi$ated more thorou$hly( or predict the
ris# level associated with vehicles at land crossin$s)
Predictive analytics enables a$encies to ma#e optimal use of
inspection staff( increase detection rates and ensure better
protection of the country and its citi"ens)
Border crossing analytics
#ne large countr with appro!imatel $%% border crossing
points uses cameras to record the registration plate of e"er
"ehicle that attempts to cross. #nce the plate is read& screens
in the crossing control booth ad"ise super"isors to either let
the "ehicle pass or direct it to the secondar inspection area.
If a "ehicle is selected for secondar inspection& information
is
sent to the mobile de"ice of the inspector showing the
likelihood of each risk tpe 'drugs& weapons& contraband& illegal
immigrants& etc.(& pro"iding the inspector with guidance on
what to look for. )ehicle selections& risk assessments and
inspection outcomes are recorded to enable ongoing reporting
on inspection rates b risk tpe& hit rates& false positi"e rates
and the amount and "alue of seizures. *his strateg makes
optimal use of inspection staff& has increased detection rates&
better protects the countr and its citizens& and impro"es the
e!perience of low+risk tra"elers b e!pediting their crossing.
,ee' communities safer
Police departments and other public safety a$encies can use
business analytics to combine data from disparate sources and
help a$encies ma#e the best use of the people and information
at hand to monitor( measure and predict crime and crime
trends) Business analytics provides insi$ht that lets officers
trac# criminal activities( predict the li#elihood of incidents
and effectively deploy resources( helpin$ to reduce crime and
increase citi"en safety and satisfaction) For instance( public
safety a$encies are now usin$ these solutions to identify
sources of crime( conduct robbery investi$ations( mana$e
repeat offenders and increase the safety of police officers)
*he &emphis Police %epartment deployed IB& Business
Analytics technolo$y to provide unparalleled insi$ht into
criminal activity and crime trends as they occurred) 3ow( the
department can chan$e tactics and redirect patrol resources as
needed to prevent crimes and catch more criminals in the act)
*he pro$ram has reduced serious crime by 1: percent( cut
violent crime by 2- percent( increased conviction rates
fourfold ? and reali"ed an 871 percent !9I( based on an
assessment conducted by independent analyst firm
3ucleus !esearch)G
*he same technolo$y can also be used by educational
institutions to ma#e their campuses safer) For instance( IB&
Business Analytics solutions can help create what5if scenarios
to determine the most effective constraints( policies and
intervention strate$ies to maximi"e the safety of students and
school personnel) 'ampus security departments can use
analytic insi$hts to anticipate the most effective ways to
allocate resources to minimi"e crime) Predictive crime insi$hts
can be delivered via email alerts( interactive dashboards and
scorecards for consistent intelli$ence that can be put to $ood
use by officers( supervisors( administrators( researchers(
professors in the criminolo$y department and others across the
institution/s ecosystem)
2: Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
As you can see( business analytics solutions enable
or$ani"ations in every industry to efficiently and reliably
analy"e the variables related to internal and external threats)
6sin$ a comprehensive but flexible set of techniCues( includin$
ris# scorin$ and anomaly detection( you can spot suspicious
conditions and react Cuic#ly to stop fraud or miti$ate the
conseCuences) @ou can also focus your investi$ative resources
on the transactions or events that are most li#ely to be
fraudulent( resultin$ in increased success rates and reduced
costs) And because the models can be updated easily(
or$ani"ations can continue to detect unusual situations or
behavior even when tactics or conditions chan$e)
Manage financial ris&
IB& Business Analytics can supplement IB&/s established ris#
mana$ement portfolio and help executives ma#e ris#5aware
decisions and meet re$ulatory reCuirements with smarter
pro$rams and methodolo$ies) *he broad ran$e of capabilities
in this area includes solutions for liCuidity ris#( mar#et ris#(
operational ris#( actuarial modelin$( $overnance( policy and
compliance mana$ement) 'ombined with predictive threat and
fraud solutions( they provide a comprehensive strate$y for
streamlinin$ ris# processes and ensurin$ an increased return on
capital in a rapidly evolvin$ mar#etplace)
Identify the ne-t%$est action
Business analytics is Cuic#ly becomin$ a pervasive technolo$y
in or$ani"ations of all types) While usa$e and adoption may
vary from the enterprise level to the departmental level( those
movin$ toward a consistent( controlled approach to mana$in$
ris#s are experiencin$ si$nificant !9I and benefits)
Some or$ani"ations are more advanced in their adoption and
use of business analytics than others( but there are $enerally
two types of approaches) Providin$ insi$ht for decision ma#ers
is the top priority for many businesses( and in such cases
advanced analytic techniCues help to paint a clear picture of
what is happenin$ and why ? re$ulatory reportin$ and ma#in$
strate$ic decisions based on #ey threat indicators( are examples
of this type of approach) 9ther or$ani"ations see# to move
beyond insi$ht to actually identify the next best action in a
mission critical process ? such as determinin$ where a police
officer should be stationed on a $iven day( whether a car should be
chec#ed at a border crossin$ or whether an insurance claim
is potentially fraudulent and reCuires further investi$ation
while it is bein$ processed)
9ur experience shows that or$ani"ations $row in analytics
maturity step by step) &any have some type of reportin$ or
analytical technolo$ies in place but then find that these
technolo$ies don/t adeCuately address critical business
challen$es or $ive the or$ani"ation full control of the
decisions they can or should ma#e) *his reali"ation moves the
or$ani"ation to ta#e steps toward becomin$ more analytically
mature throu$h the use of business analytics( which enhances
decision5ma#in$ abilities) 9r$ani"ations that ma#e this leap
are able to differentiate themselves by improvin$ processes
and proactively mana$in$ their understandin$ of and
responses to threats) We can describe these or$ani"ations as
AmastersA ? that is( they are usin$ business analytics in
sophisticated( innovative ways to protect themselves from
potential harm and usin$ the data they have to their best
advanta$e in their decision ma#in$)
IBM Software 22
"On short notice, e!re a"le to shift officers to
a #articular ard, on a #articular day, right
don to the shift level$ %t!s a "it li&e a chess
match and the %'M solution is ena"ling us to
ma&e arrests e never could have "efore$"
(Earry.odwin(Former%irectorofPoliceServices(&emphisP%
*he most advanced or$ani"ations are practicin$ information5
based decision5ma#in$ to help them mana$e threats across the
enterprise in real time) What is the next best action( based on
+
+
Infinity Property I 'asualty 'orporation of Birmin$ham(
Alabama( ma#es business decisions more accurately( more
consistently and in less time with business analytics) Since
implementin$ IB& Business Analytics solutions( Infinity has
doubled the accuracy of fraudulent claim identification(
added 02 million to its bottom line by eliminatin$ about
0
::: per month in third5party collection fees and
achieved a =:1 percent return on investment from a
reduction in claims payments and enhanced subro$ation)2:
&edeAnalytics( which helps hospitals to optimi"e the
payment collections process( incorporated IB& Business
Analytics into its solution so that clients can prioriti"e which
self5pay patients are li#ely to pay their bills( and focus their
collections efforts on this hi$h5yield se$ment of the
population)22
the information I have 4ow should I allocate my resources
And how can I ensure that threats are identified at the point of
interaction( instead of after the act has occurred What is the
ris# that this borrower will default on a potential loan
9r$ani"ations at this level are usin$ their #nowled$e of what
has happened in the past to predict what is li#ely to occur in the
future( and applyin$ that insi$ht to build strate$ies which
enable them to respond appropriately at the point of impact)
Adoption of business analytics for threat and fraud detection
can also help your or$ani"ation succeed with other critical
strate$ies that are often related( such as customer intimacy and
operational excellence) For example( when Infinity Property
I 'asualty deployed business analytics to identify potentially
fraudulent claims faster and with $reater accuracy( it also
experienced si$nificant increases in customer satisfaction)
Because the entire claims process was more efficient( le$itimate
claims submitted by low5ris# policy holders could be processed
and settled Cuic#ly( while suspect claims were fla$$ed for
special handlin$ and $iven the time and attention reCuired to
determine whether or not fraud was a factor)
Althou$h you may be at a different sta$e of adoption( achievin$
AmasterA status is an achievable tar$et if you approach it with a
solid understandin$ of how you are currently usin$ data to
control threats( and the steps you need to ta#e to reach your
ideal level)
22 Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
")ith "usiness analytics, e ere "asically
a"le to close a hole in our #oc&et here money
as lea&ing out steadily$"
(Bill%ibble(SFPof'laims9perations(InfinityPropertyI'asualty
Five steps to a proactive threat
mana$ement strate$y
In order for or$ani"ations to mana$e threats in today/s
intelli$ent( instrumented and interconnected world( they need
to loo# at how to apply business analytics at the point of
interaction( where real5time( pattern5based strate$ies mer$e
with situational context) *his level of transformation reCuires a
series of chan$es in how an enterprise mana$es information
and then applies that information to achieve its $oals)
IB& can wor# with you to create a road map that includes a
series of incremental steps desi$ned to move you towards your
threat mana$ement $oals) We recommend that before you ma#e
any decisions concernin$ specific technolo$ies or solutions( you
should be able to answer some critical Cuestions about your
or$ani"ation/s current strate$y and use of enterprise data:
2) %etermine your or$ani"ation/s current threat mana$ement
strate$y: Identify your approach to threats( both as an
enterprise and in #ey operational areas such as finance or
customer service( and the types of actions you are currently
ta#in$ to control threats) Would you describe yourself as
reactive( proactive or somewhere in between( dependin$ on
the situation or opportunity 4ow advanced would you li#e
to be in your use of business analytics 4ow are you usin$
analytics to define the parameters of your policies Is this
process dynamic( enablin$ you to modify policies as
conditions chan$e
2) ,xamine how you are usin$ business analytics today) Where
are you usin$ it( and how What types of results have you
experienced Are you accurately reco$ni"in$ threats and
opportunities( and ta#in$ preventive measures to cope
efficiently with ris#s 4ow are you proactively monitorin$
the activity that is ta#in$ place today
1) Inte$rate threat mana$ement into business as usual and
ensure the positive commitment of all sta#eholders) Are
you able to fulfill your re$ulatory obli$ations 'an you
levera$e data further so that it becomes an asset( and not Bust
a deliverable ? that is( somethin$ that can be used to drive
the or$ani"ation $oin$ forward and to minimi"e costs
=) Assess your enterprise and external data) What data are you
includin$ in your threat analysis( and what other data should
you be includin$ What types of data are available Are
departments operatin$ in silos Are you ta#in$ advanta$e of
the wealth of information that can be found in unstructured
data such as social forums( blo$s or internal sources such as
emails and call center notes
IBM Software 21
-) Identify opportunities for automation and control) Are you
proactively determinin$ which types of threats you will
allow versus those that you would li#e to prevent Are you
controllin$ the outcome when it comes to threats or is it
more of a corrective process %o you have processes or
decisions that can be easily automated 'an any decisions
be made in real time
@our responses to these Cuestions will help you be$in to
identify the areas in which you can start achievin$ real results
that will benefit your or$ani"ation)
")e ant to ma*imi+e the #roductivity of
collectors "y giving them a list of #atients ho
are more li&ely to #ay the hos#ital "ac& and
#ut the #eo#le ho are unli&ely to #ay don
at the "ottom of the list$"
(%avid&ould(Ph)%)(PredictiveAnalyticsScientist(&edeAnalytics
%rive insi$hts with IB& Business
Analytics solutions
IB& Business Analytics solutions can help or$ani"ations
achieve all the capabilities described above) IB& solutions
are easy to use( do not reCuire advanced technical s#ills and
inte$rate easily with existin$ systems and technolo$ies)
Eeadin$ or$ani"ations rely on IB& Business Analytics
solutions to help detect and prevent threats and fraud( and
#eep their personnel across their or$ani"ation fully informed
with the latest intelli$ence reportin$)
With more than =: years of analytic expertise( IB& SPSSJ
predictive analytics can help or$ani"ations predict with
confidence what will happen next so that they can ma#e
smarter decisions and improve outcomes) With the full ran$e
of statistical analysis( data and text minin$( predictive
modelin$( social media analysis and decision optimi"ation
capabilities( SPSS predictive analytics solutions can help your
or$ani"ation anticipate chan$e and $ain actionable insi$hts
from your data)
IB& 'o$nos
J
business intelli$ence and performance
mana$ement software provides or$ani"ations with the inte$rated
dashboards( scorecards( reportin$( analysis( and plannin$ and
bud$etin$ capabilities they need to $ain and act on fact5based
insi$hts) Business users are able to create detailed( customi"ed
reports without the need for I* assistance( and can drill down
into data to $ain more meanin$ful insi$hts to help them
minimi"e the ne$ative impact of the ris#s they face)
2= Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
As fully inte$rated solutions( IB& SPSS predictive analytics
and IB& 'o$nos business intelli$ence technolo$ies provide
or$ani"ations with the powerful capabilities and benefits of
business analytics) *hey can help these or$ani"ations achieve
actionable business insi$hts and superior performance(
ultimately enablin$ them to successfully execute their
strate$ies and achieve their $oals)
'onclusion
9ur world is becomin$ more instrumented( interconnected
and intelli$ent) While these factors create many opportunities for
both commercial and public sector or$ani"ations( they also
present challen$es ? such as increased exposure to threats)
*he most forward5thin#in$ or$ani"ations are turnin$ to
business analytics as a proactive approach to mana$in$ threats
because it empowers them to:
About IB& Business Analytics
IB& Business Analytics software delivers data5driven insi$hts
that help or$ani"ations wor# smarter and outperform their
peers) *his comprehensive portfolio includes solutions for
business intelli$ence( predictive analytics and decision
mana$ement( performance mana$ement( and ris# mana$ement)
Business Analytics solutions enable companies to identify and
visuali"e trends and patterns in areas( such as customer analytics(
that can have a profound effect on business performance) *hey
can compare scenarios( anticipate potential threats
and opportunities( better plan( bud$et and forecast resources(
balance ris#s a$ainst expected returns and wor# to meet
re$ulatory reCuirements) By ma#in$ analytics widely available(
or$ani"ations can ali$n tactical and strate$ic decision5ma#in$ to
achieve business $oals)
For further information please visit i$m.comHbusiness5
analytics)
+
+
+
&onitor their environments by includin$ a wide variety of
data across multiple sources)
%etect suspicious behavior to help them identify threats(
information breaches( crime and fraud)
'ontrol outcomes so that they can deliver the best response
to reduce exposure or loss and maximi"e the impact of any
action ta#en)
*e/uest a call
*o reCuest a call or to as# a Cuestion( $o to i$m.comHbusiness5
analyticsHcontactus)
An IB& representative will respond to your inCuiry within
two business days)
*hrou$h business analytics( your or$ani"ation can find a way to
$ain a deeper understandin$ of your data( and use that
understandin$ to develop proactive resolutions to the multitude of
threats and challen$es you face every day)
IBM Software 2-
K 'opyri$ht IB& 'orporation 2:21
IB& 'orporation
Software .roup
!oute 2::
Somers( 3@ 2:-8G
Produced in the 6nited States of America
&ay 2:21
IB&( the IB& lo$o( ibm)com( 'o$nos and SPSS are trademar#s of
International Business &achines 'orp)( re$istered in many Burisdictions
worldwide) 9ther product and service names mi$ht be trademar#s of
IB& or other companies) A current list of IB& trademar#s is available on
the Web at A'opyri$ht and trademar# informationA at www)ibm)comH
le$alHcopytrade)shtml)
*his document is current as of the initial date of publication and may
be chan$ed by IB& at any time) 3ot all offerin$s are available in every
country in which IB& operates)
*4, I3F9!&A*I93 I3 *4IS %9'6&,3* IS P!9FI%,%
AAS ISA WI*496* A3@ WA!!A3*@( ,LP!,SS 9! I&PEI,%(
I3'E6%I3. WI*496* A3@ WA!!A3*I,S 9F
&,!'4A3*ABIEI*@( FI*3,SS F9! A PA!*I'6EA! P6!P9S,
A3% A3@ WA!!A3*@ 9! '93%I*I93 9F 3935
I3F!I3.,&,3*)
IB& products are warranted accordin$ to the terms and conditions of
the a$reements under which they are provided)
2 Association of 'ertified Fraud ,xaminers( A2:22 .lobal Fraud StudyA
;http:HHwww)acfe)comHrttn5hi$hli$hts)aspx<
2 Ibid)
1 http:HHwww)bcbsm)comHcontentHmicrositesHhealth5care5fraudHenH
fraud5statistics)html
= Insurance Fraud( Federal Bureau of Investi$ation) ;http:HHwww)fbi)$ovH
stats5servicesHpublicationsHinsurance5fraud<
- 'redit 'ard Fraud Statistics( ;http:HHwww)statisticbrain)comHcredit5card5
fraud5statisticsH<
7 2:22 IB& .lobal !eputational !is# and I* Study( ;i$m.comHservicesHusH
$bsHbusHhtmlHris#Mstudy)html I%': *he business value of predictive analytics( Nune 2:22( ;http:HHforms)
co$nos)comHelCP6!EPa$eO=2:7IoffidOwpMspssrcMbusinessMvalueMofM
predictiveManalyticsMytw:1281<
8 Info$raphic: Bi$$est healthcare data breaches of 2:22( 4ealthcare I*
3ews) ;http:HHwww)healthcareitnews)comHnewsHinfo$raphic5bi$$est5
healthcare5data5breaches52:22<
G A!9I 'ase Study: IB& SPSS?&emphis Police %epartment)A 3ucleus
!esearch) Nune 2:2
2: AInfinity Property I 'asualty: %rivin$ the auto insurance industry
forward)A IB& 'orporation( 2:22)
22 A4elpin$ hospitals capture more revenue: &edeAnalytics employs
IB& SPSS &odeler to help prioriti"e providers/ collection efforts)A
IB& 'orporation( 2:2
Please !ecycle
@*W:12>156S,35:2
doc_949026869.docx
A smarter planet creates new possibilities, as well as new complexities and risks. Thanks to recent technological advancements, including mobile devices, cloud computing and social media, we have created an interconnected, instrumented and intelligent world in which there is virtually infinite access to information.
IBM Software Predictive Analytics Software
Business Analytics
Predictive Threat
and Fraud Analytics:
Meeting the Challenges
of a Smarter Planet
2 Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
Contents For example:
2 Introduction
What is business analytics
!educe exposure and minimi"e impact
!2 Five steps to a proactive threat and ris# strate$y
! %rive insi$hts with IB& Business Analytics solutions
!" 'onclusion
!" About IB& Business Analytics
Introduction
A smarter planet creates new possibilities( as well as new
complexities and ris#s) *han#s to recent technolo$ical
advancements( includin$ mobile devices( cloud computin$ and
social media( we have created an interconnected( instrumented
+
+
+
+
+
Accordin$ to a recent study by the Association of 'ertified
Fraud ,xaminers( the typical or$ani"ation loses - percent of
its revenue to fraud every year) .lobally( it/s estimated that
annual fraud losses exceed 01)- trillion)2
3early half of or$ani"ations that are victimi"ed by fraud are
unable to recover their losses)2
,stimates from 3ational 4eath 'are Anti5Fraud Association put
health care frauds costs in the 6S at 078 billion annually)
9ther estimates ran$e as hi$h 021: billion)1
*he total cost of insurance fraud ;non5health insurance< is
estimated at more than 0=: billion annually( costin$ the
avera$e 6S family up to 0>:: in increased premiums(
accordin$ to the FBI)=
.lobally( the total cost of credit card fraud is estimated to be
0-)- billion) In the 6S alone( ten percent of citi"ens have
been victims of credit card fraud)-
and intelli$ent world in which there is
virtually infinite access
to information) While this has opened the door to new and
excitin$ possibilities( it also provides an openin$ for new
threats and vulnerabilities) From external factors such as
national security and the economy ? to internal factors such as
insider fraud( credit ris# and information mana$ement ?
or$ani"ations face a multitude of threats every day) *hese
threats are increasin$ in number and severity( and can cost
or$ani"ations millions( even billions( in losses)
It is therefore not surprisin$ that threat and fraud miti$ation
has become a top priority for business) In addition to the
monetary losses mentioned above( there can also be a
si$nificant cost to the reputation and value of an or$ani"ation
that is unprepared for threats and fraud) Accordin$ to a recent
IB& $lobal study of senior executives( >- percent say that data
theft and cyber attac#s impact customer satisfaction and brand
reputation) And 72 percent say they are they are the $reatest
threats to their company/s reputation)7
IBM Software 1
*he best defense a$ainst threat and fraud is a systematic
approach to reducin$ exposure and minimi"in$ ne$ative
impact) But where do you be$in In an ever5chan$in$ world(
you must be dili$ent about mana$in$ the threats you are aware of(
as well as those yet to be identified) For example( you need
to be able to anticipate how an insider mi$ht infiltrate your
secure I* systems( loo# for si$ns of potential terrorist
activities( predict the li#ely impact of future economic events
or identify new patterns of fraud ? all with a hi$h de$ree of
accuracy)
In this paper we will discuss how to build a proactive threat
mana$ement strate$y based on business analytics technolo$y(
and then deploy that information to the individuals responsible
for impactin$ unwanted behavior) @ou/ll see examples of how
or$ani"ations worldwide are applyin$ business analytics
solutions to minimi"e the ne$ative impact of ris# and maximi"e
positive results) @ou will also learn the practical steps you can
ta#e to combat threat and fraud in your or$ani"ation)
What is business analytics
*he term Abusiness analyticsA refers to a comprehensive(
data5driven approach that combines the science of predictive
analytics with advanced business intelli$ence capabilities)
Predictive analytics uses advanced analytical al$orithms to
process historical data and create models that can ma#e
predictions about future outcomes) Business intelli$ence
capabilities then deliver these predictive insi$hts to #ey
personnel and departments across the or$ani"ation to help
them achieve their $oals and obBectives)
,ach of these technolo$ies is powerful in and of themselves)
When combined as business analytics( they provide superior
capabilities for improvin$ business performance across a wide
ran$e of functional areas) Accordin$ to analyst firm I%'(
or$ani"ations that utili"e predictive analytics in addition to
business intelli$ence achieve an avera$e return on investment
of 2-: percent)>
&odern business analytics software is easy to use and can be
operated by users with different s#ill levels( from front5line
business users to experienced analysts) Businesses across many
industries use business analytics to understand their customers(
increase profitability and improve operational efficiency and
effectiveness) When applied to threat and fraud prevention(
this technolo$y provides similar benefits by providin$ the
ability to connect data to effective action by drawin$ reliable
conclusions about current conditions and future events)
!educe exposure and minimi"e impact
&ountin$ re$ulatory demands( the $rowth of online
transactions and communication( the increased use of mobile
devices and the constant shadow of an uncertain economy
underscore the importance of proactively mana$in$ threats in
all their forms ? whether related to business( data or events)
*hat means not only understandin$ what has happened in the
past( but bein$ able to loo# forward to anticipate what mi$ht
happen in the future( and how your business will respond to
these situations) 9r$ani"ations that do this well can
si$nificantly maximi"e efficiency and reduce the impact of
threats( as well as differentiate their business model from those
who have not adopted a culture of analytics)
= Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
Business analytics is the #ey to reducin$ exposure and
minimi"in$ the ne$ative impact of the thousands of threats
your or$ani"ation faces every day) It provides you with access
to fact5driven predictive insi$hts in real5time( driven by your
or$ani"ation/s specific needs)
*oday/s dynamic environment reCuires a systematic approach to
mana$in$ threats) By usin$ business analytics throu$hout your
decision5ma#in$ lifecycle( you can continuously refine
the decisions you ma#e and the strate$ies you use as an
or$ani"ation to ta#e control and ta#e actions that are based on
insi$ht) As illustrated in Fi$ure 2( when you apply business
analytics( you become empowered to:
As you can see( this cycle is continuous: as you acCuire new
data( you analy"e it and learn from it) *he insi$hts you $ain
then help you improve policies and parameters so that your
or$ani"ation is constantly shielded from ne$ative ris# at
each sta$e)
%efine
Allow
+
+
+
%efine adaptive( forward5loo#in$ policy definitions and
parameters based on past actions( new challen$es and
internal or external chan$es( which can include anythin$
from the weather to evolvin$ re$ulatory reCuirements)
&onitor your environment continuously usin$ a broad ran$e
of data from multiple sourcesD addin$ new data as needed
and learnin$ from your data to develop insi$hts so that you
can identify tri$$ers and alerts)
%etect suspicious behavior such as threats( information
breaches( crime and fraud( usin$ business analytics to
Prevent
%etect
&onitor
identify anomalies and determine the li#elihood that an Figure 1: *oday/s dynamic environment reCuires a systematic approach to
mana$in$ threat and ris#)
action poses a potential ris#)
+ Prevent or allow a potential threat based on its anticipated
impact on your or$ani"ation( and either eliminate or mana$e
the ris# to ensure a controlled( output5driven approach that reduces
exposure or loss and maximi"es the positive impact
of any action ta#en) For example( a ban# can analy"e a credit
card applicant/s information( includin$ credit score( marital status(
lon$evity at current employer and current salary( to
determine how these factors into the relative ris# of default) *he ban#
may determine that while the li#elihood of default isn/t hi$h enou$h
to prevent it from securin$ the credit card(
they cannot afford to issue the card at a low interest rate( thereby
miti$atin$ some of the ris# it is assumin$)
IBM Software -
&asterin$ business analytics is a Bourney that starts with
usin$ the information you have to find insi$hts( and continues
throu$h determinin$ the next best action for a particular
#ith a $usiness analytics%$ased ris& management strategy
in 'lace you $ecome a$le to:
individual( scenario or decision ma#er) By applyin$ business
analytics techniCues or$ani"ations are able to identify #ey
threats( accommodate re$ulations( or refine and monitor
policies based on what they #now about past events as well as
data obtained in real5time)
*hat can mean( for example:
+
+
+
+
+
Access threat information across the organization in real time
to determine the best time to accept the risk or stop it
altogether.
Make decisions faster and spend less time "putting out fires."
Gain more confidence and trust in decisions that are made.
Reduce costs and make better use of resources.
Build a stronger organization that is resilient to change and
read to e!ploit new opportunities.
+
+
%eterminin$ that an auto insurance or medical claim is
fraudulent as it is processed)
Sendin$ police into areas most li#ely to have hi$h crime
activity on a particular day based on #ey predictors such as
the weather( previous days/ behavior( day of the wee# and
+
+
Support business inno"ation while reducing or maintaining
an acceptable le"el of operational risk. A"oid negati"e
publicit.
+
+
other factors)
Identifyin$ the person in your enterprise who suddenly
seems to be en$a$ed in irre$ular networ# and data activity)
Providin$ current( customi"ed threat intelli$ence reportin$
to specific employees across the or$ani"ation so they can
anticipate and address ris#s within their areas of
responsibility)
Identify organi(ational threats with
$usiness analytics
Patterns that point to threats or suspicious behavior are
often hidden in hu$e amounts of data) *his is why a proactive
approach to threat mana$ement starts with analy"in$ the
masses of data your company collects and stores such as
transactional data( email( networ# reports( survey data(
constituent information( tax records and call center notes) It/s
4avin$ the answers ahead of time $ives you the ability to
control what action you ta#e and when so that you can plan(
forecast( implement and prevent) Without analytics( you can
react only when a threat is identified( which puts your
or$ani"ation into Afi$ht or fli$htA mode( instead of enablin$
you to respond in a controlled and formulated way that will
ensure the best outcome) *his can be very costly to your
or$ani"ation ? costly in terms of lost revenue or opportunityD
reduced credibility in the eyes of employees( partners and
investorsD ne$ative environmental impactD or even lives lost)
eCually important to monitor and analy"e the massive amount
of information flowin$ throu$h social media sources such as
networ#in$ sites( blo$s and comment sections of websites)
7 Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
Business analytics solutions feature sophisticated techniCues
that enable your or$ani"ation to analy"e both the AstructuredA
data found in tables and databases as well as AunstructuredA
text( includin$ content found in email chains( chat forums(
social networ#s or survey data) In contrast to rules5based
analysis and detection methods alone( business analytics can
identify relatively unusual behaviors( even those with subtle
differences that other methods often miss) And because new
schemes continually emer$e and chan$e over time( this
process also helps identify new threats)
*his allows or$ani"ations to combine a diverse ran$e of data
types( includin$ data that is specific to your industry in terms
of benchmar#s or external data) Business analytics $ives you
the ability to combine a wide variety of data dimensions( types
and sources on an on$oin$ basis) *his ma#es it possible to
Cuic#ly and reliably detect inadvertent si$natures from hac#ers(
criminals or terrorists $eneratin$ new cyber chatter or tryin$
new tactics to $ain access to sensitive information( or to
determine the ris# of $rantin$ an individual a credit line)
%ifferent techniCues can be used to help you decide which
next best action you should ta#e)
9nce you have an understandin$ of your data( includin$ what
constitutes normal and unusual behavior( you can develop #ey
predictors( or indicators of potential threats( that can be used
alon$side your #ey performance indicators to hi$hli$ht areas
of concern for your business or or$ani"ation) @ou can then
be$in to base strate$ic and operational decisions on predictive
intelli$ence that not only shows future outcomes and behavior(
but also details the factors that influence those events) *his
increases your ability to proactively stop threats before
they occur)
Business analytics also enables or$ani"ations deliver the models
and business rules that support consistent( automated decisions
at the Apoint of impact)A *his could mean deployin$ analytic
results directly to #ey people throu$h business intelli$ence
reports( dashboards or other meanin$ful visuali"ations such as
$raphs or maps) ,mployees receive customi"ed reports that
address their particular area of concern( and are able to drill
down into the data to find root causes and $ain deeper insi$hts
for predictin$ and preventin$ the ne$ative conseCuences of
threats)
*hese insi$hts can be also embedded within operational
systems such as websites and call centers to drive smarter
decisions and actions that are ali$ned an or$ani"ation/s overall
strate$y and $oals) *his capability provides a critical layer of
optimi"ed decisions where a company/s staff or automated
systems interact with customers( prospects( suppliers or
partners) Information about results is then used to refine
predictive models and future recommended actions)
Eet/s loo# at some areas in which business analytics5based ris#
and threat control strate$ies are particularly effective)
)liminate insider threats
For most or$ani"ations( data has become an invaluable
asset ? the lifeblood of their operations) Access to this data is
available to an expandin$ user base( includin$ employees(
business partners( suppliers and customers) I* infrastructures
are more extensive( more complex( more distributed ? and
more accessible) *his interconnectedness affords many benefits
for businesses( $overnment a$encies and consumers ali#e ? but it
also introduces a $reat deal of potential ris#) 'ompanies and
$overnment a$encies must answer to strin$ent re$ulatory
reCuirements and protect intellectual capital from competitors
or subversive political entities) And consumers are typically
most concerned with the potential for identity theft and other
privacy violations)
IBM Software >
*he Wi#iEea#s scandal involvin$ the 6S State %epartment is
an ideal example of one of the primary challen$es to data
security: the authori"ed insider threat) In this instance( massive
amounts of secret documents ? includin$ thousands of
embassy cables and hundreds of thousands of documents
relatin$ to the war in Af$hanistan and IraC ? were copied by
authori"ed users to '%s and %F%s and sent to Wi#iEea#s)
Intrusion analysts at a national $overnment a$ency are often
flooded with several million events per day ? many of which
are mappin$ reCuests or information $atherin$( rather than
actual threats) *his lar$e number of events ma#es detectin$
real attac#s or potential problems difficult) 6sin$ IB&
Business Analytics software( analysts can focus on the alarms
most li#ely to be cause for concern( in one instance reducin$
the number of events reCuirin$ manual review by G> percent in
a sin$le 1:5day period)
Authori"ed insider threats are not uniCue to the $overnment or the
military) Firtually any or$ani"ation that possesses sensitive
business information such as earnin$s releases( mer$er and
acCuisition plans( mar#etin$ plans( research and development
data( personally identifiable information( or sensitive internal
emails are at ris#)
9ne of the $reatest areas of concern re$ardin$ the breach of
sensitive data is within the healthcare industry) *here have
been a number of hi$h profile cases recently in which personal
health information was stolen due to insufficient safe$uards)
For example( the 6tah %epartment of 4ealth reports that
social security numbers( addresses( dia$noses( dates of birth
and other confidential information for >8
::: patients wascompromised durin$ a hac#er attac# in 2:22) *his is not an
isolated incident) .overnment fi$ures show that hundreds of
thousands of patient records have been stolen from a variety
of healthcare or$ani"ations)8
Both the Wi#iEea#s and healthcare breaches demonstrate the
value of implementin$ business analytics technolo$y for the
prevention of data theft)
*educe credit card ris&
In today/s turbulent financial mar#et( credit card issuers need
to #now as much about their customers as possible) *hey need to
understand the factors and behaviors that indicate whether a
potential customer is a $ood credit ris#( as well as patterns or
trends that point to the probability a customer or applicant will
miss payments or default) In addition( they need to devise
more sophisticated strate$ies to #eep the $ood customers and
limit the ris#s associated with unprofitable ones) IB& Business
Analytics software allows or$ani"ations to build credit models
that are based upon a more refined se$mentation strate$y that
ta#es into account customers/ uniCue and individual data(
ensurin$ that credit policies are ali$ned with customers/ level of
ris# to ensure the most profitable outcome)
A private and independent ban#in$ institution in Swit"erland
improved the performance of its credit ris# mana$ement
system by implementin$ a new credit ris# scorin$ system based
on IB& Business Analytics software) *his solution enabled the
ban# to improve the mana$ement of credit ris# and the overall
collection procedure( and has reduced unpaid debts by 2- percent)
8 Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
+etect and 'revent fraud
Fraud is an expensive problem that costs industries and
countries billions of dollars) *he ran$e of fraud( and the
resourcefulness of fraudsters( poses a dauntin$ problem for
those char$ed with its detection and prevention) @et these
criminal activities have one thin$ in common: each of them
leaves a trail of behavioral and transactional data ? insurance
claims( mort$a$e or benefit applications( healthcare
submissions( tax returns and so on ? which are filed alon$side
information from le$itimate interactions) 'onventional
analytics can be used to detect fraud after it has happened)
IB& Business Analytics software can forecast what is li#ely to
happen in the future( as well as detect events as they happen)
'onsistent fraud intelli$ence is provided to the people on the
front lines of the business to catch fraud in the act( or even
prevent it before it ta#es place)
Minimi(e inventory loss
Business analytics helps identify products( store conditions(
personnel or customers who may be lin#ed to stoc# shrin#a$e)
Shrin#a$e solutions from IB& provide retailers with insi$hts
into product and store performance around stoc# control and
fraudulent activities( and can help increase the profitability of
store operations)
Assess networ& outages
Business analytics can help telecommunications companies or
lar$e retailers improve networ# asset mana$ement outcomes
by predictin$ which networ# will fail next( and the impact
this will have on operations and customer experience) IB&
Business Analytics solutions ensure that or$ani"ations can ta#e
Cuic# action to repair at5ris#( hi$h5traffic networ#s( enablin$
them to avoid the ris# of downtime( as well as customer
dissatisfaction and lost revenue)
+ *he anti5fraud a$ency of a southern ,uropean country
implemented IB& Business Analytics technolo$ies to Prevent energy fraud
,ner$y fraud can be difficult to detect and cause maBor
si$nificantly improve the speed( effectiveness and ease of
financial losses for ener$y a$encies and hi$her prices for
fraud detection than#s to accurate( automated identification
consumers) IB& Business Analytics solutions help prevent
of hi$h5ris# taxpayers)
ener$y fraud and theft by detectin$ suspicious activity based on
+ Eeadin$ healthcare or$ani"ations around the world use
distribution( consumption( pricin$ and smart meters and then
IB& Business Analytics solutions to minimi"e the impact of
determinin$ appropriate intervention strate$ies) *hey also
fraudulent claims by ensurin$ early detection of li#ely
help a$encies better allocate their investi$ative resources(
instances before payment occurs)
predict hi$h5ris# areas for fraud and theft( anticipate
+ 'laims and warranty departments can reduce costs and
recurrence ris#s and increase customer satisfaction
improve supplier relationships by redirectin$ fraudulent
decreasin$ the costs associated with ener$y theft)
warranty claims and presentin$ this information via
reportin$ methods inside of claims systems)
IBM Software G
Protect national $orders
Protection a$ainst threats often be$ins at border crossin$s(
airports and in harbors) IB& Business Analytics solutions help
a$encies identify which containers enterin$ a port could
contain unwantedHdan$erous materials( which passen$ers on an
airline should be investi$ated more thorou$hly( or predict the
ris# level associated with vehicles at land crossin$s)
Predictive analytics enables a$encies to ma#e optimal use of
inspection staff( increase detection rates and ensure better
protection of the country and its citi"ens)
Border crossing analytics
#ne large countr with appro!imatel $%% border crossing
points uses cameras to record the registration plate of e"er
"ehicle that attempts to cross. #nce the plate is read& screens
in the crossing control booth ad"ise super"isors to either let
the "ehicle pass or direct it to the secondar inspection area.
If a "ehicle is selected for secondar inspection& information
is
sent to the mobile de"ice of the inspector showing the
likelihood of each risk tpe 'drugs& weapons& contraband& illegal
immigrants& etc.(& pro"iding the inspector with guidance on
what to look for. )ehicle selections& risk assessments and
inspection outcomes are recorded to enable ongoing reporting
on inspection rates b risk tpe& hit rates& false positi"e rates
and the amount and "alue of seizures. *his strateg makes
optimal use of inspection staff& has increased detection rates&
better protects the countr and its citizens& and impro"es the
e!perience of low+risk tra"elers b e!pediting their crossing.
,ee' communities safer
Police departments and other public safety a$encies can use
business analytics to combine data from disparate sources and
help a$encies ma#e the best use of the people and information
at hand to monitor( measure and predict crime and crime
trends) Business analytics provides insi$ht that lets officers
trac# criminal activities( predict the li#elihood of incidents
and effectively deploy resources( helpin$ to reduce crime and
increase citi"en safety and satisfaction) For instance( public
safety a$encies are now usin$ these solutions to identify
sources of crime( conduct robbery investi$ations( mana$e
repeat offenders and increase the safety of police officers)
*he &emphis Police %epartment deployed IB& Business
Analytics technolo$y to provide unparalleled insi$ht into
criminal activity and crime trends as they occurred) 3ow( the
department can chan$e tactics and redirect patrol resources as
needed to prevent crimes and catch more criminals in the act)
*he pro$ram has reduced serious crime by 1: percent( cut
violent crime by 2- percent( increased conviction rates
fourfold ? and reali"ed an 871 percent !9I( based on an
assessment conducted by independent analyst firm
3ucleus !esearch)G
*he same technolo$y can also be used by educational
institutions to ma#e their campuses safer) For instance( IB&
Business Analytics solutions can help create what5if scenarios
to determine the most effective constraints( policies and
intervention strate$ies to maximi"e the safety of students and
school personnel) 'ampus security departments can use
analytic insi$hts to anticipate the most effective ways to
allocate resources to minimi"e crime) Predictive crime insi$hts
can be delivered via email alerts( interactive dashboards and
scorecards for consistent intelli$ence that can be put to $ood
use by officers( supervisors( administrators( researchers(
professors in the criminolo$y department and others across the
institution/s ecosystem)
2: Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
As you can see( business analytics solutions enable
or$ani"ations in every industry to efficiently and reliably
analy"e the variables related to internal and external threats)
6sin$ a comprehensive but flexible set of techniCues( includin$
ris# scorin$ and anomaly detection( you can spot suspicious
conditions and react Cuic#ly to stop fraud or miti$ate the
conseCuences) @ou can also focus your investi$ative resources
on the transactions or events that are most li#ely to be
fraudulent( resultin$ in increased success rates and reduced
costs) And because the models can be updated easily(
or$ani"ations can continue to detect unusual situations or
behavior even when tactics or conditions chan$e)
Manage financial ris&
IB& Business Analytics can supplement IB&/s established ris#
mana$ement portfolio and help executives ma#e ris#5aware
decisions and meet re$ulatory reCuirements with smarter
pro$rams and methodolo$ies) *he broad ran$e of capabilities
in this area includes solutions for liCuidity ris#( mar#et ris#(
operational ris#( actuarial modelin$( $overnance( policy and
compliance mana$ement) 'ombined with predictive threat and
fraud solutions( they provide a comprehensive strate$y for
streamlinin$ ris# processes and ensurin$ an increased return on
capital in a rapidly evolvin$ mar#etplace)
Identify the ne-t%$est action
Business analytics is Cuic#ly becomin$ a pervasive technolo$y
in or$ani"ations of all types) While usa$e and adoption may
vary from the enterprise level to the departmental level( those
movin$ toward a consistent( controlled approach to mana$in$
ris#s are experiencin$ si$nificant !9I and benefits)
Some or$ani"ations are more advanced in their adoption and
use of business analytics than others( but there are $enerally
two types of approaches) Providin$ insi$ht for decision ma#ers
is the top priority for many businesses( and in such cases
advanced analytic techniCues help to paint a clear picture of
what is happenin$ and why ? re$ulatory reportin$ and ma#in$
strate$ic decisions based on #ey threat indicators( are examples
of this type of approach) 9ther or$ani"ations see# to move
beyond insi$ht to actually identify the next best action in a
mission critical process ? such as determinin$ where a police
officer should be stationed on a $iven day( whether a car should be
chec#ed at a border crossin$ or whether an insurance claim
is potentially fraudulent and reCuires further investi$ation
while it is bein$ processed)
9ur experience shows that or$ani"ations $row in analytics
maturity step by step) &any have some type of reportin$ or
analytical technolo$ies in place but then find that these
technolo$ies don/t adeCuately address critical business
challen$es or $ive the or$ani"ation full control of the
decisions they can or should ma#e) *his reali"ation moves the
or$ani"ation to ta#e steps toward becomin$ more analytically
mature throu$h the use of business analytics( which enhances
decision5ma#in$ abilities) 9r$ani"ations that ma#e this leap
are able to differentiate themselves by improvin$ processes
and proactively mana$in$ their understandin$ of and
responses to threats) We can describe these or$ani"ations as
AmastersA ? that is( they are usin$ business analytics in
sophisticated( innovative ways to protect themselves from
potential harm and usin$ the data they have to their best
advanta$e in their decision ma#in$)
IBM Software 22
"On short notice, e!re a"le to shift officers to
a #articular ard, on a #articular day, right
don to the shift level$ %t!s a "it li&e a chess
match and the %'M solution is ena"ling us to
ma&e arrests e never could have "efore$"
(Earry.odwin(Former%irectorofPoliceServices(&emphisP%
*he most advanced or$ani"ations are practicin$ information5
based decision5ma#in$ to help them mana$e threats across the
enterprise in real time) What is the next best action( based on
+
+
Infinity Property I 'asualty 'orporation of Birmin$ham(
Alabama( ma#es business decisions more accurately( more
consistently and in less time with business analytics) Since
implementin$ IB& Business Analytics solutions( Infinity has
doubled the accuracy of fraudulent claim identification(
added 02 million to its bottom line by eliminatin$ about
0
::: per month in third5party collection fees andachieved a =:1 percent return on investment from a
reduction in claims payments and enhanced subro$ation)2:
&edeAnalytics( which helps hospitals to optimi"e the
payment collections process( incorporated IB& Business
Analytics into its solution so that clients can prioriti"e which
self5pay patients are li#ely to pay their bills( and focus their
collections efforts on this hi$h5yield se$ment of the
population)22
the information I have 4ow should I allocate my resources
And how can I ensure that threats are identified at the point of
interaction( instead of after the act has occurred What is the
ris# that this borrower will default on a potential loan
9r$ani"ations at this level are usin$ their #nowled$e of what
has happened in the past to predict what is li#ely to occur in the
future( and applyin$ that insi$ht to build strate$ies which
enable them to respond appropriately at the point of impact)
Adoption of business analytics for threat and fraud detection
can also help your or$ani"ation succeed with other critical
strate$ies that are often related( such as customer intimacy and
operational excellence) For example( when Infinity Property
I 'asualty deployed business analytics to identify potentially
fraudulent claims faster and with $reater accuracy( it also
experienced si$nificant increases in customer satisfaction)
Because the entire claims process was more efficient( le$itimate
claims submitted by low5ris# policy holders could be processed
and settled Cuic#ly( while suspect claims were fla$$ed for
special handlin$ and $iven the time and attention reCuired to
determine whether or not fraud was a factor)
Althou$h you may be at a different sta$e of adoption( achievin$
AmasterA status is an achievable tar$et if you approach it with a
solid understandin$ of how you are currently usin$ data to
control threats( and the steps you need to ta#e to reach your
ideal level)
22 Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
")ith "usiness analytics, e ere "asically
a"le to close a hole in our #oc&et here money
as lea&ing out steadily$"
(Bill%ibble(SFPof'laims9perations(InfinityPropertyI'asualty
Five steps to a proactive threat
mana$ement strate$y
In order for or$ani"ations to mana$e threats in today/s
intelli$ent( instrumented and interconnected world( they need
to loo# at how to apply business analytics at the point of
interaction( where real5time( pattern5based strate$ies mer$e
with situational context) *his level of transformation reCuires a
series of chan$es in how an enterprise mana$es information
and then applies that information to achieve its $oals)
IB& can wor# with you to create a road map that includes a
series of incremental steps desi$ned to move you towards your
threat mana$ement $oals) We recommend that before you ma#e
any decisions concernin$ specific technolo$ies or solutions( you
should be able to answer some critical Cuestions about your
or$ani"ation/s current strate$y and use of enterprise data:
2) %etermine your or$ani"ation/s current threat mana$ement
strate$y: Identify your approach to threats( both as an
enterprise and in #ey operational areas such as finance or
customer service( and the types of actions you are currently
ta#in$ to control threats) Would you describe yourself as
reactive( proactive or somewhere in between( dependin$ on
the situation or opportunity 4ow advanced would you li#e
to be in your use of business analytics 4ow are you usin$
analytics to define the parameters of your policies Is this
process dynamic( enablin$ you to modify policies as
conditions chan$e
2) ,xamine how you are usin$ business analytics today) Where
are you usin$ it( and how What types of results have you
experienced Are you accurately reco$ni"in$ threats and
opportunities( and ta#in$ preventive measures to cope
efficiently with ris#s 4ow are you proactively monitorin$
the activity that is ta#in$ place today
1) Inte$rate threat mana$ement into business as usual and
ensure the positive commitment of all sta#eholders) Are
you able to fulfill your re$ulatory obli$ations 'an you
levera$e data further so that it becomes an asset( and not Bust
a deliverable ? that is( somethin$ that can be used to drive
the or$ani"ation $oin$ forward and to minimi"e costs
=) Assess your enterprise and external data) What data are you
includin$ in your threat analysis( and what other data should
you be includin$ What types of data are available Are
departments operatin$ in silos Are you ta#in$ advanta$e of
the wealth of information that can be found in unstructured
data such as social forums( blo$s or internal sources such as
emails and call center notes
IBM Software 21
-) Identify opportunities for automation and control) Are you
proactively determinin$ which types of threats you will
allow versus those that you would li#e to prevent Are you
controllin$ the outcome when it comes to threats or is it
more of a corrective process %o you have processes or
decisions that can be easily automated 'an any decisions
be made in real time
@our responses to these Cuestions will help you be$in to
identify the areas in which you can start achievin$ real results
that will benefit your or$ani"ation)
")e ant to ma*imi+e the #roductivity of
collectors "y giving them a list of #atients ho
are more li&ely to #ay the hos#ital "ac& and
#ut the #eo#le ho are unli&ely to #ay don
at the "ottom of the list$"
(%avid&ould(Ph)%)(PredictiveAnalyticsScientist(&edeAnalytics
%rive insi$hts with IB& Business
Analytics solutions
IB& Business Analytics solutions can help or$ani"ations
achieve all the capabilities described above) IB& solutions
are easy to use( do not reCuire advanced technical s#ills and
inte$rate easily with existin$ systems and technolo$ies)
Eeadin$ or$ani"ations rely on IB& Business Analytics
solutions to help detect and prevent threats and fraud( and
#eep their personnel across their or$ani"ation fully informed
with the latest intelli$ence reportin$)
With more than =: years of analytic expertise( IB& SPSSJ
predictive analytics can help or$ani"ations predict with
confidence what will happen next so that they can ma#e
smarter decisions and improve outcomes) With the full ran$e
of statistical analysis( data and text minin$( predictive
modelin$( social media analysis and decision optimi"ation
capabilities( SPSS predictive analytics solutions can help your
or$ani"ation anticipate chan$e and $ain actionable insi$hts
from your data)
IB& 'o$nos
J
business intelli$ence and performance
mana$ement software provides or$ani"ations with the inte$rated
dashboards( scorecards( reportin$( analysis( and plannin$ and
bud$etin$ capabilities they need to $ain and act on fact5based
insi$hts) Business users are able to create detailed( customi"ed
reports without the need for I* assistance( and can drill down
into data to $ain more meanin$ful insi$hts to help them
minimi"e the ne$ative impact of the ris#s they face)
2= Predictive Threat and Fraud Analytics: Meeting the Challenges of a Smarter Planet
As fully inte$rated solutions( IB& SPSS predictive analytics
and IB& 'o$nos business intelli$ence technolo$ies provide
or$ani"ations with the powerful capabilities and benefits of
business analytics) *hey can help these or$ani"ations achieve
actionable business insi$hts and superior performance(
ultimately enablin$ them to successfully execute their
strate$ies and achieve their $oals)
'onclusion
9ur world is becomin$ more instrumented( interconnected
and intelli$ent) While these factors create many opportunities for
both commercial and public sector or$ani"ations( they also
present challen$es ? such as increased exposure to threats)
*he most forward5thin#in$ or$ani"ations are turnin$ to
business analytics as a proactive approach to mana$in$ threats
because it empowers them to:
About IB& Business Analytics
IB& Business Analytics software delivers data5driven insi$hts
that help or$ani"ations wor# smarter and outperform their
peers) *his comprehensive portfolio includes solutions for
business intelli$ence( predictive analytics and decision
mana$ement( performance mana$ement( and ris# mana$ement)
Business Analytics solutions enable companies to identify and
visuali"e trends and patterns in areas( such as customer analytics(
that can have a profound effect on business performance) *hey
can compare scenarios( anticipate potential threats
and opportunities( better plan( bud$et and forecast resources(
balance ris#s a$ainst expected returns and wor# to meet
re$ulatory reCuirements) By ma#in$ analytics widely available(
or$ani"ations can ali$n tactical and strate$ic decision5ma#in$ to
achieve business $oals)
For further information please visit i$m.comHbusiness5
analytics)
+
+
+
&onitor their environments by includin$ a wide variety of
data across multiple sources)
%etect suspicious behavior to help them identify threats(
information breaches( crime and fraud)
'ontrol outcomes so that they can deliver the best response
to reduce exposure or loss and maximi"e the impact of any
action ta#en)
*e/uest a call
*o reCuest a call or to as# a Cuestion( $o to i$m.comHbusiness5
analyticsHcontactus)
An IB& representative will respond to your inCuiry within
two business days)
*hrou$h business analytics( your or$ani"ation can find a way to
$ain a deeper understandin$ of your data( and use that
understandin$ to develop proactive resolutions to the multitude of
threats and challen$es you face every day)
IBM Software 2-
K 'opyri$ht IB& 'orporation 2:21
IB& 'orporation
Software .roup
!oute 2::
Somers( 3@ 2:-8G
Produced in the 6nited States of America
&ay 2:21
IB&( the IB& lo$o( ibm)com( 'o$nos and SPSS are trademar#s of
International Business &achines 'orp)( re$istered in many Burisdictions
worldwide) 9ther product and service names mi$ht be trademar#s of
IB& or other companies) A current list of IB& trademar#s is available on
the Web at A'opyri$ht and trademar# informationA at www)ibm)comH
le$alHcopytrade)shtml)
*his document is current as of the initial date of publication and may
be chan$ed by IB& at any time) 3ot all offerin$s are available in every
country in which IB& operates)
*4, I3F9!&A*I93 I3 *4IS %9'6&,3* IS P!9FI%,%
AAS ISA WI*496* A3@ WA!!A3*@( ,LP!,SS 9! I&PEI,%(
I3'E6%I3. WI*496* A3@ WA!!A3*I,S 9F
&,!'4A3*ABIEI*@( FI*3,SS F9! A PA!*I'6EA! P6!P9S,
A3% A3@ WA!!A3*@ 9! '93%I*I93 9F 3935
I3F!I3.,&,3*)
IB& products are warranted accordin$ to the terms and conditions of
the a$reements under which they are provided)
2 Association of 'ertified Fraud ,xaminers( A2:22 .lobal Fraud StudyA
;http:HHwww)acfe)comHrttn5hi$hli$hts)aspx<
2 Ibid)
1 http:HHwww)bcbsm)comHcontentHmicrositesHhealth5care5fraudHenH
fraud5statistics)html
= Insurance Fraud( Federal Bureau of Investi$ation) ;http:HHwww)fbi)$ovH
stats5servicesHpublicationsHinsurance5fraud<
- 'redit 'ard Fraud Statistics( ;http:HHwww)statisticbrain)comHcredit5card5
fraud5statisticsH<
7 2:22 IB& .lobal !eputational !is# and I* Study( ;i$m.comHservicesHusH
$bsHbusHhtmlHris#Mstudy)html I%': *he business value of predictive analytics( Nune 2:22( ;http:HHforms)
co$nos)comHelCP6!EPa$eO=2:7IoffidOwpMspssrcMbusinessMvalueMofM
predictiveManalyticsMytw:1281<
8 Info$raphic: Bi$$est healthcare data breaches of 2:22( 4ealthcare I*
3ews) ;http:HHwww)healthcareitnews)comHnewsHinfo$raphic5bi$$est5
healthcare5data5breaches52:22<
G A!9I 'ase Study: IB& SPSS?&emphis Police %epartment)A 3ucleus
!esearch) Nune 2:2
2: AInfinity Property I 'asualty: %rivin$ the auto insurance industry
forward)A IB& 'orporation( 2:22)
22 A4elpin$ hospitals capture more revenue: &edeAnalytics employs
IB& SPSS &odeler to help prioriti"e providers/ collection efforts)A
IB& 'orporation( 2:2
Please !ecycle
@*W:12>156S,35:2
doc_949026869.docx