Enabling the confident enterprise
with business analytics
Informed decision making • Business analytics for industries and SMBs • Analytics applied to processes • Essentials to get started
2 The impact of business analytics on performance and profitability
4 Business analytics: helping you put an informed foot forward
8 How organizations make better decisions
Thomas H. Davenport
12 Business analytics in action
Gail Bamford, David Wallace, Mike Newkirk and Becca Goren
16 The art, act and science of knowing
20 What business analytics means for small and medium businesses
23 Embedding analytics into processes
Thomas Davenport, Jeanne Harris and Robert Morison
ACCESS THIS REPORT ONLINE: 27 8 essentials of business analytics
30 The art of the possible: business analytics to
measure corporate sustainability
P2|Performance and Profitability
The impact of business analytics
on performance and profitability
By Jim Goodnight, CEO, SAS
With the rising complexity of global busi- For example, with the right analytics, re-
ness, gut decisions and hunches no tailers can predict how many red sweat-
longer suffice. Successful responses to ers they need in stock and how many
threats and opportunities now depend on smalls or larges they need based on local
rapid and smart execution. Let me state demographics. They can also determine
it plainly: Business analytics is the key to optimal prices for hundreds of thousands
achieving these challenging objectives. of products at multiple locations. Pricing
Our world generated more data in 2009 used to be an art. Now, giant retailers can
than in the previous recorded history of zero in on the optimal price for all their
mankind. A good deal of this data can SKUs and stores. Banks can determine
be converted into useful information and the optimal amount of cash to keep in
competitive advantage – by applying the ATMs. Automakers can predict how
right analytics. many spare parts they’ll need on hand –
The answers are out there – in the data
we capture and store. Harrah’s, a global casino operator, uses
analytics to optimize its marketing and
Right now, that capture and storage customer loyalty programs. Thanks
is costing huge amounts of money. largely to its use of analytics, Harrah’s
Analytics converts those tremendous ranks No. 1 in profits as a percentage
costs into invaluable assets. of revenues and has increased its share
of wallet from 36 percent in 1998 to 45
Far more than mere reporting or dash-
boards or scorecards, business analytics
is a discipline that digs deeper into these
vastly larger sets of data to uncover the
most important insights. It can mean “so-
cial network analysis” to study behaviors
and relationships on multiple levels to
uncover fraud. It can involve in-database
analytics to optimize retail assortments
or pricing. It can mean analyzing portfo-
lios to manage risk positions.
Performance and Profitability|P3
In the Philippines, the Bureau of Internal Here’s my advice: Take the time to learn
Revenue used analytics to recoup $114 about analytics. Take the time to discover
million in unpaid value-added taxes, a how analytics can provide an objective
400 percent ROI in the first year. In Swe- view of your world, not only as it appears
den, they are using analytics to reduce today but also how it’s likely to appear
the number of patients who die from clini- tomorrow. I’m not talking about gazing
cal errors. In addition to reducing unnec- into a crystal ball. I’m talking about the
essary deaths, they expect to save $10 capability of competitive organizations to
billion in health care costs at the national develop and implement strategies today
level through their analytic efforts. that are based on a careful analysis of
their likely outcomes in the future.
1-800-FLOWERS.COM changes prices
and offerings on its Web site, sometimes And here’s my crystal-ball view: The abil-
hourly, because it uses analytics. It also ity to predict future business trends with
uses analytic software to target print and reasonable accuracy will be one of the
online promotions with greater accuracy. crucial competitive advantages of this
And it uses analytics to optimize its mar- new decade. And you won’t be able to
keting, shipping, distribution and manu- do that without analytics.
facturing operations. The result: a $50
million reduction in costs last year.
Business Analytics Knowledge Exchange
Jim Goodnight has been at SAS’ helm since the
company’s incorporation in 1976, overseeing an
unbroken chain of revenue growth – a feat almost
unheard of in the software industry.
P4|Face Forward with Business Analytics
Business analytics: helping you
put an informed foot forward
By Jim Davis
Most companies today have plenty of Why BI is not enough
data. Creating intelligence and glean- Business intelligence provides histori-
ing real insight from this data is what cal, metric-driven decision making –
continues to elude organizations. De- and answers questions like, how many
spite years of talk about scorecards and units did we sell, what did customers
metrics, gut feelings and experience are buy and for how much? BI is charac-
often still the guides for making impor- terized by the creation of simple rules
tant, sometimes critical decisions, even and alerts and the distribution of known
though current research reveals a clear facts to systems and people. These
link between business performance and decisions have a low transformational
the use of business analytics. impact on the business.
So what exactly is business analytics BI is still a highly valuable part of your
and how can it help? Business analytics overall business analytics environment,
is, simply put, the application of ana- however, offering an excellent general
lytical techniques to resolve business purpose backbone for ad hoc analysis
issues. It provides organizations with a and basic operational reporting.
framework for decision making, helping
organizations solve complex business For example, BI can alert management
problems, improve performance, drive on how many credit card transactions
sustainable growth through innovation, were completed on a given day. It can
anticipate and plan for change while also develop a simple rule for automatic
managing and balancing risk. reporting, like reporting on transactions
greater than $10,000 to the regulators.
It sounds like a lot, but if you break it
down it’s all about enabling effective From a more strategic decision perspec-
decision making. Organizations make tive, business analytics can help answer
decisions every day, and these sit on a questions such as what new products
continuum from frequent, up to millions should we offer and in what markets?
per day to transformative, which occur Or relative to the example, which credit
less frequently but greatly impact orga- card transactions are likely to be fraudu-
nizational strategy. The need for agile lent? Business analytics can predict this
decision making has never been greater with certainty and automatically deny
but unfortunately, IT infrastructure, peo- transactions – while reporting activities
ple and processes are lagging behind. in real time.
Face Forward with Business Analytics|P5
A business analytics
framework is not a
monolithic and costly
approach but rather
provides for incremental
growth to achieve
strategic goals at any
given stage of an
organization’s value chain.
Business analytics allows organizations In this way, business analytics drives In the following report, you’ll hear from
to “face forward,” bringing insight to innovation and improves an organiza- several experts about how business
transformative decisions. It benefits all tion’s speed of response to market and analytics can be applied to business
aspects of an organization’s value chain, environmental changes. In the credit card problems across all types of organizations,
including: scenario, business analytics can not industries and value chains. Perhaps
• Inbound logistics: receiving, storing, only discover the causal factors of fraud, then it will become part of your plan to
inventory control and transportation but also forecast accurately when it will outthink and out-smart the competition.
scheduling. occur again. The company can then
change business processes accordingly.
• Operations: including factors such as
packaging, equipment maintenance, A step toward business analytics
testing and all activities that add value Effective decision making requires
from the raw material to final product. a business analytics framework that
• Outbound logistics: the activities re- incorporates the people, processes,
quired to get the finished products to technology and culture of an organiza-
market, including warehousing and tion. This common framework provides
distribution management. flexibility across the entire range of
analytical decision-making types from
• Marketing and sales: activities that highly managed operational analytics
lead a buyer to purchase the product, (such as a setting a simple credit limit)
including channel selection, advertis- to discovery-based analytics (such as
ing, promotion, selling, pricing, retail credit fraud scenarios or setting dynamic
management and shelf space optimi- credit limits).
A business analytics framework is not
• Service: activities that maintain a a monolithic and costly approach,
product’s value, including customer but rather provides for incremental
support, repairs, installation, training, growth to achieve strategic goals at any
spare parts management and more. 1
given stage of an organization’s value
chain. It offers business-ready analytical ONLINE
applications with underlying technolo- Business Analytics Knowledge Exchange
gies for key services like data man- www.sas.com/baexchange
Porter, Michael E., Competitive Advantage : Creating
agement and quality, reporting and Credit card fraud management
and Sustaining Superior Performance. 1985. advanced analytics. www.sas.com/ba-cardfraud
P6|Face Forward with Business Analytics
Six questions about your company’s information
The modern organization is awash in information – yet, too often, it falls short of the
tools, methods and expertise it needs to derive the greatest value from this untapped
asset. Information about the most important facets of the business – customers, processes,
employees, competitors and more – is gathered but not analyzed, reported but not
understood, guessed about rather than acted upon. But not with business analytics.
Ask these questions of your company and join aggressive competitors by being a smart
1. Where should we leverage business analytics? Focus business analytics where you
already compete. The payoff is greatest where you are playing to your strength, not where
you are playing catch-up.
2. Why now? Because the technology is ready. Because competitors are likely exploring
the possibilities of analytical competition, too. And because it’s always risky to delay
capitalizing on a new business capability.
3. What’s the payoff? Business analytics is all about anticipating the payoff in order to
maximize it. The analytics initiative succeeds when the business capitalizes on an
opportunity that analytics reveals.
4. What information and technology do we need? Most companies don’t lack for sufficient
data, but instead suffer from a lack of integration and a lack of quality. Without good data,
you simply can’t do good analytics.
5. What kind of people do we need? You need a variety of talented people: analytical
professionals who design and refine analytical algorithms, and perform data mining;
analytical semiprofessionals who do substantial amounts of modeling and analysis but
are unlikely to develop sophisticated new algorithms or models; analytical amateurs who
need to understand something of the analytical basis for operations and decisions; and
the analytical manager who focuses the work of analytical professionals.
6. What roles must senior executives play? Committed senior executives provide the passion
and the resources to drive their organizations in an analytical direction. In virtually every
successful firm, senior management sets an analytical strategy and continually pushes
Face Forward with Business Analytics|P7
HSBC: fraud detection that exceeds
With fraud levels surging around the
world, banks are facing greater regula-
tory scrutiny, as well as risks associated
with damaging publicity from fraud. The
ability to correctly make split-second
decisions on accepting credit card
transactions – before fraud occurs – is
more important than ever.
Using SAS Fraud Management, part of
the SAS Business Analytics Framework,
HSBC prevents, detects and manages
financial crimes by scoring and accept-
ing or rejecting millions of transactions a
day in real time – at the point of sale.
As a result, the global financial services
leader has achieved significantly lower
incidence of fraud across tens of mil-
lions of debit and credit card accounts.
“The proof is in our fraud numbers – our
detection rates and our false positives –
which continue to meet our aggres-
sive goals,” said Derek Wylde, Head of
Group Fraud Risk, Global Security and
Fraud Risk for HSBC.
Jim Davis is Senior Vice President and
Chief Marketing Officer for SAS.
How organizations make
The following article is an edited excerpt of an article distributed by the
International Institute for Analytics.
By Thomas H. Davenport
Relatively few businesses and organiza-
tions have given full and proper attention
to one of their most important activities:
making decisions regarding key questions
such as what strategies and business
models to pursue, which products and
services to offer, which customers to
target, what prices to charge and what
Author and researcher Tom Davenport is the employees to hire. Organizations with
President’s Distinguished Professor at Babson College. poor decision processes and tools
eventually encounter poor outcomes,
His newest book is Analytics at Work: Smarter Decisions, and performance suffers.
Better Results (with Jeanne Harris and Robert Morison,
However, new analytics, decision auto-
from Harvard Business Press). mation tools and business intelligence
systems make it possible to make better
use of information in decisions. “Wisdom
of crowds” approaches and technologies
allow larger groups of people to partici-
pate meaningfully in decision processes.
Organizations cannot afford to ignore
these new options if they wish to make
the best possible decisions.
Given both negative and positive incen-
tives to get better, one might expect
that organizations would attempt to
improve their decisions — that they
would prioritize them, examine their
current level of effectiveness, investigate
new options for making them better and
implement some of those options. In
my survey and analysis of dozens of
corporations, I found that while they are,
indeed, doing some of these things,
In a survey and analysis
of dozens of corporations,
Davenport found that very
few organizations have
efforts to improve a variety
very few organizations have undertaken decision-oriented context from the start.
systematic efforts to improve a variety of If a test between two alternative Web
decisions. In this excerpt I describe some page designs is performed, it is gen-
of the more frequent approaches used to erally assumed that a decision to adopt
intervene in decision processes. the winning page will be made. Other
analytical approaches may not have as
Analytics, testing and data clear a path to a decision.
Infrastructures predicated on analytics
and data were among the most A prerequisite of virtually any form of
common decision-making frameworks analytics is high-quality data, so it is not
among the surveyed firms. Eighty-four surprising that data-oriented responses
percent of respondents mentioned an were also common. Sixty-six percent
analytical component in their decision of respondents mentioned some issue
improvement efforts and 66 percent involving data. The most common were:
mentioned efforts to improve data.
The range of analytical techniques • Having difficulty in accessing data.
employed was quite broad. Scoring
• Creating a common data architecture.
approaches based on statistical analyses
(usually some form of regression analy- • Eliminating duplicate data.
sis) were common. Other approaches
included optimization, behavior-based • Integrating “master data
customer targeting, statistical forecasting, management.”
prediction of various phenomena and the
• Achieving “one version of the truth”
use of text analytics.
in functional or process areas. Not surprisingly, many
Systematic testing was one form of
• Dealing with too much data.
analysis that was being used somewhat
that they needed to change
frequently by companies; 18 percent • Gathering data from channel partners.
mentioned it specifically in interviews. business processes to
One key virtue is that it creates a • Creating new metrics. make better decisions.
Technology support – and overrides Changes in business processes • An insurance company adopted
– for decisions Not surprisingly, many organizations enterprise risk management.
Several firms surveyed mentioned spe- reported that they needed to change
cific analytical software, testing software, business processes to make better • The Six Sigma approach to process
data warehouses and Web analytics/ decisions. Forty-three percent men- quality and decision outcomes was
reporting software. Two other tech- tioned process changes of some type. implemented at a financial payments
nologies were mentioned frequently: For instance, some described process firm and a staffing firm.
specialized information display technolo- changes around supply chain manage-
• A financial services firm uses the
gies and business rule engines. ment in an IT firm, lease processing in
“net promoter score” for customer
an auto financing firm, financial process-
Thirty-eight percent of companies in the satisfaction decisions.
es in health insurance or new product
study mentioned some use of specialized development processes. Several organi- • An economic decision analysis
information displays such as scorecards zations mentioned changes for decision- approach, popularized and taught
and dashboards. These tools, typically oriented processes made in the context by Stanford’s Engineering School
found in the business intelligence of Six Sigma programs. and the Strategic Decisions Group,
category, allow decision makers to see
is used by an oil company.
only the information that they need to However, some decision-focused ana-
make a decision. Several firms mentioned lysts noted that their original goal wasn’t In addition, three responding organiza-
using specific display approaches not necessarily to identify and implement tions developed analytically focused
generally supported by conventional BI process changes, and that they had to decision processes that have been widely
tools, including the “A3” format for work with other groups to accomplish used in IT systems development, but are
displaying key issues in a particular them. As one head of an analyst group not widely known in the decision-making
business domain. Some companies are at an IT firm commented, “We didn’t or analytics literature. Sometimes called
using neuroscience principles to guide initially have the franchise to do process “agile methods” or “rapid prototyping,”
how information is presented and improvement — our thing was analytics. they involve the creation of a series of
digested. This may be a bellwether of But it kept coming up on our projects. So short-term deliverables, and frequent
future attempts to link information and we eventually just made it a part of our review of them by the client and stake-
decision making. standard approach.” holders for the decision. The organi-
zations that use this approach found
Another popular decision technology Decision-oriented methods and tools
that it led to results that better fit the
involves using business rules to enable Several organizations reported that
decision-makers’ requirements, and at
automated or semiautomated decision one aspect of their decision processes
a faster pace.
processes — sometimes in conjunction was an overarching, strategic manage-
with analytics (e.g., scoring-oriented ment approach to guide all aspects of
applications). Many organizations em- their efforts. Most of these initiatives are
ploy business rules but allow humans well-known approaches to business
to override the recommended decisions and management.
From my research, it’s clear that
organizations recognize the importance Analytics improves
of improving decisions. Although the decisions
survey was not a random sample, Davenport’s research found the most
individuals in 90 percent of organiza- common types of decisions improved by
tions surveyed identified some analytics include:
attempt to improve decisions through • Pricing decisions (consumer goods,
better processes. Second, organiza- industrial goods, government contracts,
tions employ a variety of intervention maintenance contracts, etc.).
types to improve decisions across
• Decisions to target consumer segments
analytics, culture and leadership, and
(by retailers, insurers, credit card firms).
data. The most successful organiza-
tions adopted multiple interventions • Merchandising decisions (brands
to buy, quantities and allocations).
at once to improve a decision.
• Location decisions (for bank
As a result, analysts — previously branches or where to service industrial
responsible for data gathering and equipment). ONLINE
analysis — are morphing into consultants • Treatment protocols for health care. Order it now – Analytics at Work: Smarter
who may be responsible for framing deci- Decisions, Better Results
• Product development for http://www.analyticsatworkbook.com/
sions, process redesign, communication
and education programs, and change Read the full International Institute for
management — all in addition to the • Student performance in educational
traditional analysis functions.
• Evaluating marketing approaches
Engage with analytic leaders
Organizations seeking to implement (in both consumer and
decision improvements should become B2B environments). www.iianalytics.com
familiar with these common intervention • Hiring decisions.
types and create ongoing capabilities to
• Vehicle routing decisions.
Analysts — previously responsible for data
gathering and analysis — are morphing into
consultants who may be responsible for framing
decisions, process redesign, communication and
education programs, and change management —
all in addition to the traditional analysis functions.
P12|Business Analytics in Action
Business analytics in action
How are key industries deriving value from their business analytics implementations?
By Gail Bamford, David Wallace, Mike Newkirk and Becca Goren
HEALTH CARE • Increased compliance on medication
According to the World Health Organi- reconciliation by more than 50 percent
zation, global health spending totalled in a nine-month period.
more than US$4.1 trillion in 2007, with
$639 as the total health expenditure • Dramatically reduced the rate of hos-
per person. That number will only grow pital-acquired infections by measuring
in ways that affect businesses and where infections originated and what
citizens. admission conditions closely corre-
lated with acquired infections.
Despite these huge investments, health
care quality is uneven and resistant to • Improved government/industry ac-
changes and improvements. How can creditation/compliance by incorpo-
rating national guidelines into key
we enhance health care delivery while
controlling those costs? It starts by
carefully measuring and monitoring the • Developed new methods for caring for
quality of that care – a complex task stroke patients while controlling costs.
perfectly suited for business analyt- By taking better care of these patients,
ics. Here’s how some forward-thinking the hospital expects fewer complica-
health care institutions are delivering tions, which will reduce costs.
better quality of care more efficiently.
Maine Medical Center The Karolinska Institute in Sweden
Named to US News and World Report’s needed a way to examine the effects
“America’s Best Hospitals” list for of drugs, other treatments and lifestyle
orthopedics, heart care and gynecologic factors on patients with rheumatoid ar-
care, Maine Medical Center uses SAS thritis. Using SAS Business Analytics,
Business Analytics to understand key the Institute has deployed a Web-based
patient care metrics – and sustain a patient self-help application and predic-
quality-driven culture. The data-driven tive modeling to determine which treat-
approach has produced excellent results: ments will be most effective for certain
segments of RA patients.
Business Analytics in Action|P13
BANKING • Cleanse and integrate. Cleanse and
In a challenging economic and regula- standardize third-party credit and
tory climate, bankers must be especially customer data, enrich it (e.g., add
vigilant. Two key indicators of a bank’s geocoding tags) and integrate it into
health are net charge-offs (NCOs) – the a single data store.
value of loans written off as uncollect-
able – and nonperforming loans (NPLs) • Analyze and score. Develop scoring
that are in default or delinquent more models to analyze debtor-customer
than 90 days. segment data against objectives, in-
cluding “maximize profits” or “minimize
In the past two years in the US, bank writeoffs” or against constraints, such Optimizing collections
NCOs have soared by an average of as loan types, outstanding balances or A leading Australian financial institution
more than 350 percent across all insti- days delinquent. previously relied on instinct when contact-
tutions, with institutions holding assets ing delinquent customers. Since introduc-
of $5 billion or less showing growth of • Optimize and execute treatment ing SAS for collections optimization, it has
almost 500 percent. NPLs as a percent- strategies. Analytical models help achieved a 300 percent ROI in less than six
age of average loan balances have risen collections teams understand who is months. A debt purchasing firm based in
more than 278 percent at US banks with most likely to respond, which commu- the UK uses SAS to predict debt portfolio
$1 billion or more in assets. How can fi-
nication channels work best and how performance. This enables the firm to
nancial institutions improve their collec- much payment to expect. make quicker decisions on acquiring new
tions and protect their bottom line? debt portfolios at the right prices, collect
Collections optimization driven by more from each portfolio and grow rev-
Business analytics can provide the in- business analytics delivers the results enues by 50 percent annually.
sights that institutions need to reduce that institutions need to improve their
both loan writeoffs and the cost of col- profitability.
lections activities. First, models created
within a business analytics framework
can identify likely candidates for work- 1
Source: SNL Financial
outs and loan modifications. Second,
business analytics can optimize collec-
tions activities to improve the probability
of success and maximize self-treatment
among debtor segments. It starts with
three basic steps.
P14|Business Analytics in Action
MANUFACTURING such as point-of-sale (POS) data and
From diapers to jet engines and almost historical shipment data. Once that data
everything in between, manufacturing is aggregated, business analytics models
expertise is a competitive differentiator for and tools can accurately forecast the
companies that follow optimal practices demand for products by family, individual
and methodologies to attack inefficiencies SKU, geography, customer type, etc.
and eliminate waste. Business analytics With a clear and accurate demand
is essential in these settings to improve picture, manufacturers can properly
production and sales planning, enhance allocate raw materials across plants and
the supply chain, reduce inventory, regions – all optimized by distribution
streamline logistics and much more. channel – to create complete roll-ups in
Meaningful ROI with master planning schedules.
Business Analytics For example, with demand forecasting,
One SAS customer increased company business analytics can be a key TELECOMMUNICATIONS
profitability by accurately predicting prod- contributor to a manufacturer’s success. You’ve likely experienced it before – your
uct demand and customer behavior – more Better forecasts deliver ROI by: cell phone loses service one too many
than doubling its forecasting accuracy. It times, so you switch providers. Low
found that for every 1 percent reduction in • Reducing inventories. barriers to churning mean providers must
forecast variance, it saved $200,000. vigilantly and carefully invest to maintain
• Improving order fulfillment rates. and increase their service quality and
Another manufacturer improved two
seemingly competing objectives. It simul- customer satisfaction rankings. After
• Shortening cash-to-cash cycles.
taneously reduced inventory by 20 percent, all, your satisfaction keeps them in
eliminating millions of dollars of holding Many manufacturers struggle with business.
costs, yet improved service levels, which optimally managing and forecasting
directly and positively affected customer Network managers typically receive error
their raw materials requirements, work-
satisfaction. reports and alarms after a network device
in-process (WIP) inventory and finished
fails. The team addresses the stream of
goods inventories. Without the right mix of
trouble tickets, but never gets insight into
raw materials, production plans fall apart
underlying causes or trends for outages.
and customer orders are delayed (or,
The result: long call-resolution times.
worse, canceled). Missing WIP forecasts
similarly leads to inefficient schedules With business analytics and approaches
and a crippling misallocation of finished such as predictive fault analysis, network
stocks – not having the right quantities of managers can analyze performance
the right goods at the right time and in to pre-empt failures. They can analyze
the right places. While the data is often trouble tickets and optimize corrective
available to prevent, identify and correct services, shortening times you are
these imbalances and inefficiencies, it without coverage.
is usually not integrated, analyzed and
shared across the organization. Strong data management, including
data quality and reporting capabilities –
Data management technologies can all key underpinnings for business
bring together islands of information analytics – can help quickly identify
Business Analytics in Action|P15
One large telco service provider used
SAS to identify emerging issues (an
average of two weeks prior to failure)
and double the percentage of tickets
resolved within 48 hours.
service and network issues. Business
analytics helps to:
• Identify and remove duplicate trouble
• Understand faults and performance on
a macro level.
• Determine which services have the
highest fault rates.
In addition to analyzing network
performance, predictive analytics
technologies can help evaluate
demand, faults and systems to improve
resource utilization and quality of service
(QoS). A telco provider can then identify Gail Bamford is a SAS Global Industry Marketing
Manager for Public Sector.
when and where network resources
are deployed and quality/performance
variations over time. David Wallace is a SAS Global Industry Marketing
Manager for Financial Services.
Business analytics allows network and
service managers to better understand Mike Newkirk is a SAS Global Industry Marketing
Manager for Manufacturing.
causes and impacts of failures. They can
prioritize and pre-empt outages, optimize
repairs and mitigate risk with answers to Becca Goren is a SAS Global Industry
Marketing Manager for Communications,
key questions: Media and Entertainment.
• How significant is each factor
influencing network faults degradation? ONLINE
Health care providers keep pace with change
• Which network faults are tied to a given www.sas.com/ba-healthcareprovider
trouble ticket? The standard for clinical data analysis and reporting
• Which faults are related and what are
their impacts? Solutions for better risk management
Armed with predictive fault analytics, Get the full stories on:
a telco provider can limit the times you Maine Medical Center Compete in manufacturing
lose a signal and continually improve
overall service, allowing it to keep your Karolinska Institute Invest wisely, communications service providers
business. www.sas.com/ba-karolinska www.sas.com/ba-telco
P16|The New Know
The art, act and science of knowing
An excerpt from The New Know 1
By Thornton May
The Internet makes self teaching — and
lifelong learning — the rule rather than
the exception. Historians ultimately will
come to consensus on what to call the
time period between the frenzy that was
the dot-com bubble and the period be-
fore society finally enters the data cloud.
For want of a better phrase, I call the 20-
year interregnum we currently inhabit
(1995 – 2015) the Age of Little Informa-
tion. I come to this label not because the
age exhibits a lack of information. Quite
the contrary, it is during this epoch that
information — previously locked away
in analog form — is becoming widely
digitized. The New Know has changed
our reality along 10 fundamental dimen-
Futurist Thornton May positions analysts as heroes New Know Reality #1:
of the age we are about to enter in his new book, You will be expected to do
The New Know: Innovation Powered by Analytics. something with information.
All this newly digitized information has
had, relatively speaking, little impact on
behavior and little impact on organiza-
tional outcomes. We are now exiting a
historical moment of undermanaged and
only occasionally acted-upon informa-
tion to an environment requiring much
more active, much more intense, much
more aggressive information manage-
ment. You as an executive will be held
much more accountable for your data
management behaviors. You will be
expected to transform “data lead” into
Copyright 2009 by John Wiley & Sons, Inc.
All rights reserved. Reprinted with permission.
“knowledge gold” via the expeditious
The New Know|P17
sensemaking leading to efficacious ac- New Know Reality #3:
tion. In the Age of Little Information, we You will have to know more
were data vegetarians. In the New Know about knowing.
we will have to become information and
knowledge carnivores. One of the major changes defining the
new competitive environment is the
New Know Reality #2: requirement to know more about know-
There really is more to know. ing, what experts sometimes refer to as
metacognition. Society is about to
The New Know will be awash with undergo a tectonic shift in how it thinks
data. Processing power doubles every about thinking. Driving this cognitive
18 months. Storage capacity doubles plate shifting are the RSS feeds, pod-
every 12 months. Bandwidth through- casts, blogs, old-media headlines and
put doubles every nine months. There evening news programs, which are
is more to know. Organizations are increasingly filled with images and
having trouble keeping up — and, sadly, instances of current-generation leaders
the fact that there are more facts arriving being asked by dissatisfied next-
at a faster rate of speed is not even the generation voters, customers and
tip of the cognitive iceberg. Like the “fog shareholders: “What were you
of war,” info warriors speak of the “fog thinking?” Looking beneath the surface,
of facts” (e.g., confusion about what they are really asking: “How were you
information is to be believed, what infor- thinking? Via what processes, using
mation sources are credible and what what data and assisted by what tools did
version of reality is to be acted on). In a you arrive at your course of action?”
world of multiple sources of information
and 24-hour decision making, the very New Know Reality #4:
character of information is changing. Brain science and decision
A “fact” is no longer a “fact.” science are converging.
Scientists do not know how the brain
works — yet. But they are sneaking up
on it. Readers may be surprised to learn
that neuroscience has been around
for over 100 years. Neuroscience has
progressed to the point that we at least
know what we do not know.
P18|The New Know
Carly Fiorina, former
CEO at Hewlett-Packard,
believes that distilling
truth from overwhelming
amounts of information is To some extent, it is a simple truism New Know Reality #6:
the essence of leadership. that the brain is involved with all things Information management Is the
that comprise our human existence. essence of leadership.
It follows, loosely, therefore, that
understanding the brain will help us Low-cost communications give rise to
understand the human condition more almost toxic levels of spin, hype and
fully. The big news is that the brain pos- empty rhetoric. Leaders are able to
sesses innate qualities that influence cut through all the noise. Does your
individual experience and opinions. organization filter its data? Carly Fiorina,
There are things that can be known— former CEO at Hewlett-Packard,
that need to be known by executives believes that distilling truth from over-
seeking to maximize value from the whelming amounts of information is the
knowledge assets available to the essence of leadership. She believes that
enterprise. all of us are overwhelmed with informa-
tion, and what sets great leaders apart
New Know Reality #5: is their ability to cut through the clutter
The environment is changing and distinguish the truly important from
our brain. the merely interesting.
The information flood should be viewed New Know Reality #7:
as a permanent macroenvironmental A more connected world.
change. Thinking in Darwinian terms,
what adaptive pressures does this One of the transformational elements
environmental change place on us? moving society to the New Know is
“Daily exposure to high technology — something analysts at Forrester Re-
computers, smart phones, video games, search call the “groundswell.” Josh
search engines — stimulates brain cell Bernoff, Vice President at Forrester,
alteration and neurotransmitter release, contends: “There’s so much information
gradually strengthening new neural flowing out of the groundswell, it’s like
pathways in our brains while weaken- watching a thousand television chan-
ing old ones. Because of the current nels at once. To make sense of it, you
technological revolution, our brains are need to apply some technology, boil-
evolving right now — at a speed like ing down the chatter to a manageable
never before.” stream of insights.” The new scarce re-
source in the next economy will be the
human attention needed to make sense
of information. The question is: How will
we be able to keep up?
The New Know|P19
New Know Reality #8: useful to know where it will step. Every
Math matters. key process in your enterprise is locked
in a room with an elephant — a critical
Mathematics is now so widely accept- process, serving a critical customer.
ed as the arbiter of truth in the modern Business analytics tells you where that
world that it has become the backbone elephant will step.
of disciplines ranging from physics (of
course) to economics and sociology. New Know Reality #10: If you are locked in a
Backing up a statement with mathemat- Knowing can change the world.
ics gives it an aura of validity, even if the room with an elephant, it
topic has to do with something as math- If knowledge is power, then “knowl- is useful to know where
edge about power should be especially
ematically messy as human behavior.
empowering,” says John Murrell, the
it will step. Every key
However, many otherwise “normal” ex- very-much-in-the-know editor of Good process in your enterprise
ecutives have a pathological aversion Morning Silicon Valley. For instance, is locked in a room with
to math. This is not just unfortunate, it using 15,000 meters, a subset of Na-
an elephant — a critical
is dysfunctional. Some intuition about tional Grid Customers will be able to
numbers, counting and mathematical access their energy — use information process, serving a critical
ability is basic to almost all animals. via the Internet, by a thermostat read- customer. Business
People use math to make decisions out, or through text messaging, and use analytics tells you where
every day. “In an age where you need the data to change their consumption
to be numerate to do almost anything patterns. Program participants are ex- that elephant will step.
(from building bridges to conquering pected to save 5 percent, or about $70
disease), governments anxiously com- a year, on their energy bills. Change ad-
pare their performance in mathematics vocates from all fields of endeavor are
with that of competitor nations.” excited about the possibility of putting
new information in front of people in the
New Know Reality #9: hopes of changing behavior.
There are significant downsides to
Success requires materially expanding
what you know and adding precision
and efficiency to the processes (analyt-
ics) whereby you come to know. Here
is a metaphor to keep in mind as you
think about the New Know. If you are
locked in a room with an elephant, it is
P20|Business Analytics for SMBs
What business analytics means for
small and medium businesses
An interview with Matthew Mikell, SMB Global Product Marketing Manager
When it comes to business analytics, general constraints when listening to
it sometimes seems like only major en- organizations that are SMBs:
terprises garner the spotlight. That’s
somewhat understandable given the 1) Decision-making style
complexity and scope of their analytical Transitioning from gut instinct to fact-
The Wine House discovers challenges and the nature of their high- based framework can be difficult in part
$400,000 in ‘lost’ inventory profile brands. But the fact is, far more because the former approach has likely
small to medium businesses (SMBs) are served the successful SMB very well.
Economic times may be tough, but Bill
poised to implement business analytics Most SMBs have Excel experts who
Knight, owner and President of The Wine
House, is toasting a 100 percent return on solutions. can generate some great static charts
his investment in SAS. The first day its SAS and graphs — and I wouldn’t ever want
In the US, these companies have to denigrate the value those reports
application was live, the brick-and-mortar
revenues of less than $500 million. In provide. But there’s so much more val-
and Internet retailer discovered 1,000
items of wine that hadn’t moved in more Europe, the SMB category comprises ue that can be derived from in-depth
than a year. companies with a maximum of EUR 450 analyses. Once SMB executives get
million (about US$611 million). While in a real glimpse of the insights that are
“We had a huge sale to blow it out, gener- the Asia Pacific region, SMB often refers lurking beneath the surface of their
ating $400,000 in capital in one weekend,” to both employee numbers and revenue, transaction data, their willingness
Knight said, “and just in time, because in and range between 200 and 250 to adopt business analytics increases
today’s economy, we’d be choking on that employees and $200 million and $500
inventory.” pretty quickly.
million in revenue. In many ways, these
Using SAS, The Wine House has reduced businesses are striving for the same 2) Cash flow
its aged inventory by 40 percent. “Now I goals to grow their business through
In addition to a shift in decision-mak-
can get the answers I need and base de- innovation, and need the same sophisti-
ing style, cash constraints can pose
cisions on facts rather than gut intuition,” cated functionality scaled appropriately
very real obstacles for an SMB that
says Knight. “I’ve got less money tied up to their processes. In this Q&A,
wants to mature in this area. Consid-
in inventory, I know who our best custom- Matthew Mikell, SAS Global Product
ering the business analytics frame-
ers are, how to market to them and can Marketing Manager, shares his perspec-
work helps improve margins, retain key
monitor the effectiveness of our marketing. tives on what business analytics means
Our ROI with SAS has been well over 100 customers and grow share of wallet in their
percent in less than a year, so my return on markets. However, the long-lasting
investment has been fantastic.” Q. What are some of the unique return on investment far outweighs the
challenges that SMBs face with capital required to undergo the transition.
ONLINE respect to business analytics?
A: SMBs primarily face the issue of
scale. At SAS we have heard four
Business Analytics for SMBs|P21
SMB executives – often
owners or people with
lengthy tenures – worry
about letting go of the
information flow and
empowering people to
make decisions that
were previously reserved
3) IT resources and infrastructure at a disadvantage. Internally, employees
More than 80 percent of SMBs with need these tools to be productive. Oth-
about 100 employees have only four erwise, it’s gut-based decisions, or cut-
dedicated IT staffers. They’re stretched ting and pasting from multiple tools.
thin, and that can make it very difficult The truth is what brought you to where
to expand the IT mandate beyond criti- you are typically won’t take you to the
cal business operations into managing next level. But it’s very difficult, cultur-
business analytics environments. ally, to walk away from what’s made
4) Business analytics maturity you successful. SMB executives – often
owners or people with lengthy tenures –
SMBs must have an appreciation for worry about letting go of the information
the level of skills required to meet over- flow and empowering people to make
all strategic goals through business decisions that were previously reserved
analytics. Research from Aberdeen for executives.
Group suggests that SMBs without the
relevant skill sets are poorly positioned Q What’s the best way for SMBs
to drive value from an analytical solu- to tackle the adoption of business
tion. It reports that SMBs using some analytics?
sort of analytical applications perform
at a higher level than their competitors A: Of course, every company differs –
that do not.1 particularly at the SMB size. But we’ve
found that there is a general approach to
The main SMB challenge for moving to the adoption of business analytics. The
business analytics is the understanding first step is to ensure you have sponsor-
of its impact on these four critical areas, ship from company executives. Clearly
and building a capability that is cost- lay out the business analytics benefits
effective and remains flexible and easy and return to the management team.
to use. This transparency is key at the SMB
level as SMB executives are tradition-
Q. Why should SMBs adopt ally heavily involved in analyses, report-
business analytics? ing and the decision-making process.
Make it clear how business analytics will
A: It essentially boils down to competi-
resolve a compelling issue or attract and
tive pressures. SMBs need to continu-
retain customers, for example.
ally innovate. If you’re an SMB that isn’t
constantly seeking to optimize every
possible aspect of the operation, you’re 1
Aberdeen Group, 2009, Beyond Spreadsheets:
The Value of BI and Analytics.
P22|Business Analytics for SMBs
The second strategy is to focus on a Q. What’s the difference between Q. Can you share some examples
particular business process or issue. business analytics for large enter- of how SMBs have been able to
Don’t introduce business analytics as prises vs. SMBs? capitalize on business analytics?
a broad, unfocused utility for general
usage. This will occur naturally as you A: In a nutshell, it’s about scale. Deploy- A: Sure. We’ve worked with an energy-
solve more focused issues, building up ment and support strategies will have a trading company that enables staff to
confidence in fact-based decision mak- different nature. What’s more interesting predict what today’s electricity and
ing as a core competency. Some of the to me, however, is the important com- gas purchases will sell for months later
typical issues that we see being solved monality: functionality. Business analyt- when consumers buy. Business analytics
with business analytics include improv- ics in SMBs is not about presenting a supplies that intelligence to traders in a
ing customer data quality for improved subset of functionality but rather surfac- cleaner, faster and more accurate way.
marketing, invoicing or customer ser- ing the right functionality for the problem
at hand, and opening up to more as the A collection agency uses SAS Business
vice, or improved product pricing and
business requires it. Despite their size, Analytics to analyze bad-debt portfolios
packaging analysis to drive a higher
SMBs face similar challenges to make before acquiring those assets. This is a
better and more informed decisions to quantum leap forward from its previous
Finally, don’t rest on your laurels. Capi- continue innovating in their markets. It is model, which was simply buying any
talize on your initial success to broaden therefore essential to provide a rich set debt assets for as little as possible and
deployment to other areas of the orga- of features and a very high level of tech- hoping to collect successfully.
nization. Those adoptions move faster nology usability.
once you can point to a successful track A player in the secondary-ticket market
record in another area. uses SAS to develop a deeper under-
standing of the needs of its thousands
of customers. By segmenting them and
catering to psychographics, the company
can optimize how frequently it contacts
the customers and improves loyalty.
Software for SMBs
Some of the typical issues
that we see being solved
with business analytics
customer data quality
for improved marketing,
invoicing or customer
service, or improved Matthew Mikell leads Global Product Market-
product pricing and pack- ing for SMB markets and software-as-a-service
(SaaS) offerings at SAS, supporting strategic
aging analysis to drive a planning, messaging and product offerings
higher market share. through direct and indirect channels.
Analytics at Work|P23
Embedding analytics into processes
In their latest book, Analytics at Work: Smarter Decisions, Better Results,
Thomas Davenport, Jeanne Harris and Robert Morison show how companies
apply analytics in their daily operations. This excerpt, ‘Embedded Analytics in Action,’
explores what to consider when infusing analytics into business processes.
We see examples of analytics at work The solutions naturally demand very
within core processes in a variety of sophisticated and industrialized ana-
business areas. Statistical analysis has lytics: for capacity planning of aircraft
been a feature of supply chain and lo- and truck fleets, for routing packages
gistics management for decades, start- through its distribution network, and for
ing with the techniques of statistical scheduling and routing delivery trucks.
process control (SPC) and total quality For a company this steeped in analyti-
management (TQM). cal applications, the frontier is moving
closer to real-time, dynamic adjustments.
Real-time analytics are helping guide For example, UPS is experimenting with
call center workers in their interactions algorithms to adjust the order of deliv-
with customers. And analytics are well eries as conditions (e.g., road closures,
established in the engineering and sim- extraordinary customer need) change.
ulation sides of product design.
Making processes analytical
Among business support functions, The effects of analytics on the opera-
analytics are essential to many facets tions of a process can be profound,
of finance, common in the management and over time you may want to reengi-
of technology operations, and rela- neer the overall business process and
tively new to human resources (though revamp its information systems to
of enormous potential there). In cor- capitalize on the potential for analyt-
porate development, key decisions — ics-based improvement. But you can
for example, regarding mergers and start embedding analytics without a
acquisitions—may benefit greatly from major overhaul. For processes that rely
analytics, but few companies take a extensively on enterprise systems, it
process approach to such activities. may be possible to simply start taking
advantage of the analytical capabilities
Consider the example of UPS to whet
that are already included in the soft-
your appetite for embedding analytics
ware. However, many process analytics
in your core business processes. As a
initiatives will require tools, techniques,
logistics company, UPS lives and
and working relationships that are likely
breathes the “traveling salesman
to be new and unfamiliar at first. We
problem”—how to reach a variable
have found that implementing analytics-
Reprinted by permission of Harvard Business Press. Excerpted series of destinations most efficiently
from Analytics at Work: Smarter Decisions, Better Results by
enabled processes requires applying
with the right delivery capacity, and often
Thomas Davenport, Jeanne Harris and Robert Morison. four major perspectives.
All rights reserved. in designated time windows, every day.
P24|Analytics at Work
The effects of analytics The first is process implementation. Third is systems implementation. The
on the operations of a Occasionally a business may create analytical system must be incorporated
process can be profound, a new analytically enabled process into the set of systems and technolo-
and over time you may or rebuild a process from scratch, but gies supporting the business process.
most often you are adding capability to In building these interfaces, it helps to
want to reengineer the and altering an existing process. Espe- employ process-oriented technologies,
overall business process cially given the iterative nature of many including capabilities of ERP systems,
... but you can start analytical applications, it’s essential to workflow and document management
measure baseline process performance systems. And integrating and testing
embedding analytics first and to run the enhanced process the new systems and interfaces is criti-
without a major overhaul. in parallel to the original (perhaps as a cal given analytics’ reliance on a broad
pilot or test) in order to refine the new range of quality data and the fact that
process and measure its performance analytics-based decisions may dramat-
and value. In some cases, process ically change process flow.
simulation can yield insights about how
the process might perform even before Human implementation is the fourth
implementation. perspective. Often the greatest imple-
mentation challenge, especially when
Next, organizations should consider analytics is new to the process and the
model implementation. Much of the people performing it, is on the human
distinctive work of process analytics side. Only people can tell if an embed-
centers on designing, developing and ded application is resulting in good
iteratively refining statistical algorithms decisions, so be sure to involve them in
and descriptive or predictive models developing, managing and monitor-
or rule-based systems. If you are go- ing the assumptions and results of any
ing to industrialize important decision embedded model. Another important
processes, it is important that the rules, factor is developing the right mix of
assumptions and algorithms in your automated and human decision making
model are correct. Analytical projects and enabling process performers to
generally require different tools and trust and use their new analytical infor-
development methodologies from mation and sometimes tools.
those employed in more traditional sys-
tems development. And, of course, this
work is performed by business analysts
and programmers with special skills in
statistical methods and modeling.