The Deciding Factor: How big data is transforming business decision-making

Business Analytics The way we see it

The Deciding Factor:
Big Data & Decision Making

Written by

The Deciding Factor: Big data and decision-making

Foreword

Big Data represents a fundamental shift in business decision- The survey also highlights special challenges for decision-
making. Organisations are accustomed to analysing internal making arising from Big Data; although 85% of respondents
data – sales, shipments, inventory. Now they are increasingly felt the issue was not so much volume as the need to analyse
analysing external data too, gaining new insights into and act on Big Data in real-time. Familiar challenges relating
customers, markets, supply chains and operations: the to data quality, governance and consistency also remain
perspective that Capgemini calls the “outside-in view”. We relevant, with 56% of respondents citing organisational silos
believe it is Big Data and the outside-in view that will generate as their biggest problem in making better use of Big Data.
the biggest opportunities for differentiation over the next five For our respondents, data is now the fourth factor of
to ten years. production, as essential as land, labour and capital. It follows
that tomorrow’s winners will be the organisations that succeed
The topic of Big Data has been rising rapidly up our in exploiting Big Data, for example by applying advanced
clients’ agenda, and Capgemini is already undertaking predictive analytic techniques in real time.
extensive work in this area all over the world. That is why we
commissioned this survey from the Economist Intelligence I would like to thank the teams at the Economist Intelligence
Unit: we wanted to find out more about how organisations are Unit and within Capgemini, along with all the survey
using Big Data today, where and how it is making a difference, respondents and interviewees. I believe this research will do
and how it will be used in the future. much to increase understanding the business impact of Big
Data and its value to decision-makers.
The results show that organisations have already seen
clear evidence of the benefits Big Data can deliver. Survey
participants estimate that, for processes where Big Data Paul Nannetti
analytics has been applied, on average, they have seen a 26%
improvement in performance over the past three years, and Global Sales and Portfolio Director
they expect it will improve by 41% over the next three.

2


About the Research
43%
Capgemini commissioned the The Economist Intelligence Unit
Economist Intelligence Unit to write The conducted a survey, completed in
Deciding Factor: Big data and decision- February 2012, of 607 executives.
making. Participants hailed from across the
globe, with 38% based in Europe, 28%
The report is based on the following in North America, 25% in Asia-Pacific
research activities: and the remainder coming from Latin
America and the Middle East and of participants are C-level
Africa. The sample was senior, 43% of and board executives
participants being C-level and board
executives and the balance—other
high-level managers such as vice-
presidents, business unit heads and
department heads. Respondents
worked in a variety of different functions
and hailed from over 20 industries.
Of the latter, the best represented
were financial services, professional
services, technology, manufacturing,
healthcare and pharmaceuticals,
and consumers goods and retail.

To supplement the survey, the
Economist Intelligence Unit conducted
a programme of interviews with
senior executives of organisations
as well as independent experts
on data and decision-making.

Sincere thanks go to the survey
participants and interviewees for
sharing their valuable time and insights.

3


Executive summary
When it comes to making business At the same time, practitioners query unstructured data, such as text
decisions, it is difficult to exaggerate interviewed for the report—all analytics and sentiment analysis. A
the value of managers’ experience enthusiastic about the potential large number of executives protest that
and intuition, especially when hard for big data to improve decision- unstructured content in big data is too
data is not at hand. Today, however, making—caution that responsibility difficult to interpret.
when petabytes of information for certain types of decisions, even
are freely available, it would be operational ones, will always need Although unstructured data
foolhardy to make a decision to rest with a human being.
without attempting to draw some
causes unease, social media
meaningful inferences from the data. Other findings from the research are growing in importance.
include the following:
Anecdotal and other evidence is Social media tell companies not
only what consumers like but, more
indeed growing that the intensive use The majority of executives importantly, also what they don’t
of data in decision-making can lead
to better decisions and improved
believe their organisations like. They are often used as an early
business performance. One academic to be “data driven”, warning system to alert firms when
study cited in this report found that, but doubts persist. customers are turning against them.
controlling for other variables, firms Forty-three percent of respondents
that emphasise decision-making based Fully two-thirds of survey respondents agree that using social media to make
on data and analytics have performed say that the collection and analysis of decisions is increasingly important.
5-6% better—as measured by output data underpins their firm’s business For consumer goods and retail,
and performance—than firms that strategy and day-to-day decision- manufacturing, and healthcare and
rely on intuition and experience for making. The proportion of executives pharmaceuticals firms, social media
decision-making. Although that study who say their firm is data-driven is provide the second most valued
examined “the direct connection higher in the energy and natural datasets after business activity data.
between data-driven decision-making resources (76%), financial services
and firm performance”, it did not (73%), and healthcare, pharmaceuticals The job of automating
and biotechnology sectors (75%).
question the size of the data-sets
They may not be as data-savvy as
decision-making is
used in decision-making. In fact, very
their executives think, however: far from over.
little has been written about the use
of “big data”—which is distinguished majorities also believe that big data
management is not viewed strategically Automation has come a long way, but a
as much by its large volume as by majority of surveyed executives (62%)
the variety of media which generate at their firm, and that they do not have
enough of a “big data culture”. believe there are many more types
it—for decision-making. This report is of operational and tactical decisions
an attempt to address that shortfall. that are yet to be automated. This
Organisations struggle is particularly true of heavy industry
The research confirms a growing to make effective use where regulation and technology have
appetite among organisations for data
of unstructured data held automation back. There is, to be
and data-driven decisions, despite their
struggles with the enormous volumes for decision-making. sure, a limit to the decisions that can be
automated. Although technical limits
being generated. Just over half of are constantly being overcome, the
executives surveyed for the report say Notwithstanding the heavy volumes, increasing demand for accountability—
that management decisions based one-half of executives say they do especially following the financial
purely on intuition or experience are not have enough structured data to crisis—means that important business
increasingly regarded as suspect, and support decision-making, compared decisions must ultimately rest with a
two-thirds insist that management with only 28% who say the same about human, not a machine. For less critical
decisions are increasingly based on unstructured data. In fact, 40% of or risky decisions, however, there is still
“hard analytic information”. Nine in respondents complain that they have much scope for decision-automation.
ten of the executives polled feel that too much unstructured data. Most
the decisions they’ve made in the past business people are familiar with
three years would have been better if spreadsheets and relational databases,
they’d had all the relevant data to hand. but less familiar with the tools used to

4


This is particularly true of machine-
to-machine communication, where
low-risk decisions, such as whether to
replenish a vending machine or not, will
increasingly be made without human
intervention.

Organisational silos
and a dearth of data
specialists are the main
obstacles to putting big
data to work effectively
for decision-making.
Data silos are a perennial problem,
and one which the business process
reengineering revolution of the
1990s failed to resolve. Regulation
and the emergence of “trusted data
aggregators” may help to break down
today’s application silos, however.
Arguably a longer term challenge is
the lack of skilled analysts. Technology
firms are working with universities to
help train tomorrow’s data specialists,
but it is unlikely that supply will
meet demand soon. In the near
future, there is likely to be a “war for
talent” as firms try and outbid each
other for top-flight data analysts.

5


Introduction
26%
Moneyball: The Art of Winning an Unfair resources. Although financial services
Game, by Michael Lewis, is the story of and healthcare firms have long been
an underperforming American baseball big data users—where big data is
team—the Oakland Athletics—that defined by its enormous volume
turned a losing streak into a winning and the great diversity of media
streak by intensively using statistics and which generate it—heavy industry
analytics. According to the New York appears to be catching up (see case
is the extent of performance Times, the book turned many business study: GE—the industrial Internet).
improvement already people into “empirical evangelists”1.
experienced from big data. Nine in ten survey respondents agree
An Economist Intelligence Unit survey, that data is now an essential factor of

41%
supported by Capgemini, of 607 senior production, alongside land, labour and
executives conducted for this report capital. They are also optimistic about
found that there is indeed a growing the benefits of big data. On average,
appetite for fact-based decision- survey participants say that big data
making in organisations. The majority has improved their organisations’
of respondents to the survey (54%) say performance in the past three years
that management decisions based by 26%, and they are optimistic that
purely on intuition or experience are it will improve performance by an
is the performance increasingly regarded as suspect (this average of 41% in the next three
improvement expected view is held even more firmly in the years. While “performance” in this
in the next three years. manufacturing, energy and government instance is not rigorously specified,

55%
sectors), and 65% assert that more it is a useful gauge of mood.
and more, management decisions are
based on “hard analytic information”. One may question whether the
surveyed firms are as “data-driven”
Until recently there was scant research as their executives say. The research
to back the Moneyball hypothesis—that also shows that organisations are
if organisations relied on analytics for struggling with the enormous volumes
decision-making they could outperform of data and often with poor quality
say that big data their competitors. In 2011, however, data, and many are struggling to free
management is not viewed Erik Brynjolfsson, an economist at the data from organisational silos. The
Sloan School of Management at the same share of respondents who say
strategically at senior levels Massachusetts Institute of Technology their firms are data-driven also say
of their organisation. (MIT), along with other colleagues there is not enough of a “big data
studied 179 large publicly traded culture” in their organisation; almost
firms and found that, controlling for as many – 55% – say that big data
other variables, such as information management is not viewed strategically
technology (IT) investment, labour and at senior levels of their organisation.
capital, firms that emphasise decision-
making based on data and analytics When it comes to integrating big data
performed 5-6% better—as measured with executive decision-making, there
by output and performance—than is clearly a long road to travel before
those that rely on intuition and the results match the optimism. This
experience for decision-making2. report will examine how far down that
1
www.nytimes.com/2011/10/02/business/after- road firms in different industries and
moneyball-data-guys-are-triumphant.html Two-thirds of the executives in the regions are, and will shed light on the
survey describe their firm as “data- steps some organisations are taking to
2
Brynjolfsson, Erik, Hitt, Lorin M. and Kim, Heekyung driven”. That figure rises to 73% make big data a critical success factor
Hellen, “Strength in Numbers: How Does Data-Driven for respondents from the financial in the decision-making process.
Decision making Affect Firm Performance?” (April 22, services sector, 75% from healthcare,
2011). Available at SSRN: http://ssrn.com/abstract=1819486 pharmaceuticals and biotechnology,
or http://dx.doi.org/10.2139/ssrn.1819486 and 76% from energy and natural

6


On average, respondents believe that big data will improve organisational
performance by 41% over the next three years

Survey Question: Approximately to what extent do you believe that the use of big data has improved your
organisation’s overall performance already, and can improve overall performance in the next three years?

Now 3 Years

45%

40%

35%

30%

25%

20%

15%

10%

5%

Average CEO/President CFO/Treasurer CIO/CTO

7


Overall, 55% of respondents state that they feel big data management is not viewed
strategically at senior levels of their organisation

Survey Question: To what extent do you agree with the following statement:

“Big data management is not viewed strategically at senior levels of the organisation.”

Strongly Agree Agree Disagree Strongly Disagree Don’t know/Not applicable

100%

80%

60%

40%

20%

0% Total Financial Energy & Consumer Health & Manufacturing
Sector Resources Pharmacy

Two thirds of executives believe that there is not enough of a “big data culture” in
their organisation - this is particularly notable across the manufacturing sector

“There is not enough of a “big data culture” in the organisation, where the use of big data in decision-making is
valued and rewarded.”


100%

80%

60%

40%

20%

0% Total Financial Energy & Consumer Health & Manufacturing
Sector Resources Pharmacy

8


Putting big data
to big use
“A lot of people will say data is To keep customers loyal, retailers
important to their business, but I think have to target customers with
it’s incredibly important to healthcare personalised loyalty bonuses,
and it’s probably getting more and discounts and promotions. Today, most
more important,” says Lori Beer large supermarkets micro-segment
executive vice president of executive customers in real time and offer highly
enterprise services at WellPoint, an targeted promotions at the point of
American healthcare insurer. Ms Beer sale.
compares data in healthcare with
“oxygen”—without it, the organisation
would die.
Business activity data and point-of-sale data are
WellPoint has 34 million members, and considered most valuable across the consumer
making sure their customers get the goods & retail sector
right diagnosis and receive the right
treatment is vital for keeping costs
under control. But getting to the right
Survey Question: Which types of big data sets do you see as adding the most
information to make the right decision
value to your organisation?
in healthcare is no mean feat. There
are terabytes to sift through: millions [select up to three options]
of medical research papers, patient
records, population statistics and
Total Consumer goods & retail Top 3
formularies, to name a few types of
needed information. Using that to make
an effective decision requires powerful 68.7% 32.0% 27.7% 25.2% 21.9% 18.6% 15.5% 15.5% 10.2% 8.1% 4.3%
computing and powerful analytics (see
WellPoint case study). 57.9% 7.9% 42.1% 71.1% 18.4% 21.1% 13.2% 10.5% 5.3% 7.9% 0.0%

There is near consensus across
industries as to which big data sets
are most valuable. Fully 69% of survey
respondents agree that “business
activity data” (eg, sales, purchases,
costs) adds the greatest value to
their organisation.The only notable
exception is consumer goods and retail
where point-of-sale data is deemed to
be the most important (cited by 71% of
respondents). Retailers and consumer
Business activity data

Office documentation
(emails, document stores)

Social media

Point-of-sale

Website clickstream data

Website clickstream data

Geospatial data

Telecommunications data
(eg phone or data traffic)

Telemetry - detailed activity
data from plant/equipment

Images / graphics

Something not on this list
(please specify)

goods firms are arguably under more
pressure than other industries to
keep their prices competitive. With
smartphone apps such as RedLaser and
Amazon’s Price Check, customers can
scan a product’s barcode in-store and
immediately find out if the product is
available elsewhere for less.

10


42%
Office documentation (emails, media to express their anger at the
document stores, etc) is the second charge. Verizon Wireless was prompt
most valued data set overall, favoured in responding to the outcry, possibly
by 32% of respondents. Of the forestalling customer defection to rival
other major industries represented mobile operators.
in the survey, only healthcare,
pharmaceuticals and biotechnology But not all unstructured data is as easy
of survey respondents say differ on their second choice. Here to understand as social media. Indeed,
that unstructured content is social media are viewed as the second 42% of survey respondents say that
too difficult to interpret. most valuable data set, possibly unstructured content—which includes
because reputation is vitally important audio, video, emails and web pages—is
in this sector, and “sentiment analysis” too difficult to interpret.
of social media is a quick way to identify A possible reason for this is that today’s
shifting views towards drugs and other business intelligence tools are good at
healthcare products. aggregating and analysing structured
data whilst tools for unstructured data
Over 40% of respondents agree that are predominantly targeted at providing
using social media data for decision- access to individual documents (eg
making has become increasingly search and content management).
important, possibly because they It may be a while before the more
have made organisations vulnerable advanced unstructured data tools, such
to “brand damage”. Social media are as text analytics and sentiment analysis,
often used as an early warning system which can aggregate and summarise
to alert firms when customers are unstructured content, become mass
turning against them. In December market. This may be why 40% of
2011 it took Verizon Wireless just one respondents say they have too much
day to make the decision to withdraw unstructured data to support decision-
a $2 “convenience charge” for paying making, as opposed to just 7% who feel
bills with a smartphone, following a they have too much structured data.
social media-led consumer backlash.
Customers used Twitter and other social

40% of respondents believe that they have too much unstructured data to support
decision-making

Survey Question: Looking specifically at your department, how would you characterise
the amount of data available to support decision-making?

Too much Enough Not enough Don’t know

Structured Unstructured

7.0% 42.1% 49.8% 1.2% 39.6% 30.8% 27.6% 2.0%

11


Enough data or too much?
Structured or unstructured, most
executives feel they don’t have enough
data to support their decision-making.
In fact, 40% of respondents overall
Case study: Big data at the bedside
believe the decisions they have made For WellPoint, one of America’s In January 2012, WellPoint began
in the past three years would have been largest health insurers, the problem training the supercomputer for the
“significantly better” if they’d had all of of ensuring the right treatment plan is first phase of the project. The pilot
the structured and unstructured data provided for its members is becoming system helps WellPoint nurses review
they needed to make their decision. increasingly complex. “Getting and authorise treatment requests from
And, despite the fact that respondents relevant information at the point- medical providers. It is an iterative
from the financial services and energy of-care, when decisions are getting process where the nurses follow
sectors are more likely than average to made, is the holy grail,” says Lori Beer, the existing procedures, examine
describe their firm as data-driven, they executive vice president of enterprise the response the system provides,
are also more likely than the average business services at WellPoint. and then score it based on how well
(46% from financial services, and 48% it does. The feedback is used to
from energy) to feel they could have By some estimates, the body of educate and fine-tune the system
made better decisions if the needed medical knowledge doubles every so that it will eventually be able to
data was to hand. five years. Coupled with an explosion authorise treatments without human
in medical research papers is the intervention.
At first blush, this may seem rapid conversion of medical records For the second phase, WellPoint
contradictory, given the surfeit of data to electronic format. A physician has has partnered with Cedars-Sinai
and the difficulty organisations face in a pile of digital information to sift Samuel Oschin Comprehensive
managing it, but Bill Ruh, vice president, through yet, according to Ms Beer, Cancer Institute in Los Angeles to
software, at GE sees no contradiction. most healthcare providers spend develop a decision-support system
“Because the problems we address are very little time with each patient and for oncologists. It is hoped that
going to get more and more complex, only see “a slice of the information”. physicians will be able to review
we’re going to solve more complex WellPoint wants to provide all the treatment options suggested by the
problems as a result,” he says. “What we relevant information that a healthcare supercomputer at the point of care.
find is the more data we have, the more provider needs, in digestible format, Critically, the system won’t just provide
we get innovation in those analytics and at the patient’s bedside. an answer; it will show the oncologist
we begin to do things we didn’t think we the documented medical evidence
could do.” “If you look at the statistics, evidence- that supports the probability of why it
based medicine is only applied about believes the answer is accurate.
For Mr Ruh, the journey to data 50% of the time,” says Ms Beer. “The
fulfilment will be over when he can put issue we often face is that we’re “It is the physician who makes the
a sensor on every component GE sells not really using the most relevant ultimate decision,” says Ms Beer. “This
and monitor the component in real time. evidence-based medicine in diagnosis is not intended to ever replace the
In this way, any aberrant behaviour can and treatment decisions.” A wrong physician.”
be immediately identified and either diagnosis and treatment plan can be
corrected through a control mechanism deadly for a patient and very costly for There is no end date for the project,
(decision automation) or through human WellPoint. and various decision-support and
intervention (decision support). “We’re decision-automation tools will be
really trying to get to what we would call WellPoint had been following developed over time. The intent is
‘zero unplanned outages’ on everything the advances of IBM’s Watson that the more the WellPoint system is
we sell,” says Mr Ruh. supercomputer for some time and trained, the more accurate diagnoses
realised that the natural-language- and treatment plans will become. If
processing abilities of the machine this pans out, it will help to drive down
would make it ideal for processing the cost of healthcare in the US, where
petabytes of unstructured medical wasted health spending in 2009 was
information, and drawing meaningful estimated to be between $600 billion
conclusions from it in seconds. and $850 billion.

12


The virtues & risks
of automation
58%
Data can either support a manager in With corporate clients, however, it is
making a decision (eg, information on much more difficult. “Suppose that
key performance indicators displayed a ship cannot leave a port due to
on a business intelligence “dashboard”) late payment, and suddenly all the
or it can automate decision-making bananas go rotten; from a commercial
(eg, an automatic stock replenishment perspective, this involves a much higher
algorithm). According to the survey, on risk because the amounts are much
average big data is used for decision larger,” says Mr Knorr. “The human on average use big data
support 58% of the time, and 29% of the element and review by somebody for for decision support.

29%
time it is used for decision automation. larger amounts of money won’t go
For Michael Knorr, head of integration away.”
and data services at Citi, a financial
services group, deciding whether to However, the job of automating
use big data for decision support or decision-making at Citi is far from
decision automation depends on the over. Mr Knorr says the drive for more
level of risk. automation comes from the increasing
expectations of customers and
“In the consumer space, where amounts regulators for rapid decision-making. of the time it is used for
are small and if you make an error it’s “If you do not have the right level of decision automation.
easy to compensate for that error, then automation in place, that means your
automation might be applicable,” says costs have increased,” says Mr Knorr. “If
Mr Knorr. If there is a “false positive”— there is more data and you haven’t kept
that is, a loan is rejected by the system up with automating, then the number of
based on various set parameters when items you need to review manually will
it should have been approved—the have increased, which means you need
situation can easily be remedied with a more resources and people to do so.
phone call. This strengthens the business case for
automation.”

14


60% of respondents dispute the proposition that most operational/ tactical
decisions that can be automated, have been automated

“Most operational/tactical decisions that can be automated, have been automated.”

54.1%

48.2%
45.2%

34.2%

27.7%
25.1%

13.9% 15.2%

8.2% 8.9%
6.6%
3.6% 3.9% 3.4%
1.7%

North America Europe Asia–Pacific

Latin America
Middle East & Africa

5.1% 5.1%

16.9% Total
5.0% Strongly Agree
29.1% Agree
35.6%
37.3% 48.7% Disagree
13.5% Strongly Disagree
3.8% Don’t know

15


Across all industries and regions, a
majority of survey respondents concur
that there is scope for further decision
automation at their firm. Over 60% of Case study: General Electric and
respondents dispute the proposition
that “most operational/tactical
the industrial Internet
decisions that can be automated, have If the first phase of the Internet was electric vehicle charging stations.
been automated.” This view is fairly about connecting people, says Bill
consistent across industries, although Ruh, vice president of software at “We are putting more and more
fewer healthcare and pharmaceuticals General Electric (GE), then the second sensors on all the equipment that we
companies agree with the statement phase is about connecting machines. sell, so that we can remotely monitor
(52%) than manufacturing companies Some people call this “the Internet of and diagnose each device,” says
(68%). (Respondents from the education things”, but Mr Ruh prefers the term Mr Ruh. “This represents a huge
sector also appear less certain than “the industrial Internet”. Like many productivity gain, because you used
peers elsewhere that there is much good ideas, the concept preceded to require a physical presence to know
still to be automated.) There is some the technology. But now, sensors and what was going on. Now we can sell a
regional variation, too. No more than big data analytics have reached a level gas turbine and remotely monitor its
54% of executives in Asia-Pacific believe of maturity that makes the industrial operating state and help to optimise
the job of automation is incomplete, Internet achievable. Machines are it.”
compared with 71% in western Europe. able to talk to each other over vast
distances and make decisions without “Trip Optimizer” is a fuel-saving
Mr Ruh of GE explains why automation is human intervention. system that GE has developed for
far from complete in his industry: “One freight trains. It takes into account
reason is that many of the environments “When you look at business process a wealth of data, including track
we operate in are highly regulated, so automation, the main productivity conditions, weather, the speed of the
we have to move at a speed that makes gains have been the low hanging train, GPS data and “train physics”,
sense within the regulation,” he says. fruit in the consumer, retail and and makes decisions about how
“The second is because the sensors entertainment sectors,” says Mr and when the train should brake. In
and the data weren’t really there to Ruh. “But we have not seen many tests, Trip Optimizer reduced fuel
automate anything.” automation and productivity gains use by 4-14%, according to Mr Ruh.
in industrial operations.” National With fuel being one of the biggest
Certainly decision-automation tools electricity grids, for example, are some overheads for freight train companies
have evolved from simple “if then of the world’s biggest “machines”, (at Canadian Pacific, one user of GE’s
else” programmable statements (eg, yet the fundamentals around how system, it makes up nearly one-quarter
“if credit rating = AAA, then approve the technology is used and how it of operating costs), a 10% reduction in
loan, else reject”) to sophisticated interacts with other systems have not fuel use represents a huge cost saving.
artificial intelligence programs that kept pace over the course of a century.
learn from successes and failures. But with sensors, control systems and Mr Ruh likens the industrial Internet
The more sophisticated the tools the Internet, a “smart grid” could to Facebook or Twitter for machines.
become, the more decisions that can make decisions, such as which energy Whether it is a jet engine or oil rig,
be automated. Decision automation, supply to switch to, or which part of a machine is constantly providing
however, can introduce unnecessary the network to isolate in the case of a status updates on performance. Big
rigidity into business processes. At fluctuation or disturbance. data analytics look for patterns in
times of high instability—such as the performance, and when an anomaly
current economic climate—companies In November 2011, GE showed its is identified, a decision about the
need to be nimble in order to adapt to commitment to catching up with the best corrective action is automatically
the changing conditions. Hard-coded business-to- consumer (B2C) sectors taken or a person is alerted so that
decisions can be costly and time- by opening a new software centre in a decision can be made on the best
consuming to change. San Ramon, California, with Mr Ruh as course of action.
its head. GE is in the process of hiring
400 software engineers (with 100 on “I believe that we’re in the early stages
board to date) to complement the of this,” says Mr Ruh, “and we haven’t
company’s 5,000 software workers even begun to imagine the algorithms
who are focused on developing we’re going to build and how they’re
applications for power plants, going to improve the kinds of products
aeroplanes, medical systems and and services we offer.”
Brynjolfsson, Erik, "Riding the Rising Information Wave–

Are you swamped or swimming?", MIT Sloan Experts,

http://mitsloanexperts.com/ /2011/05/18.

16


Standing in the way
The perceived benefits of The road to these riches, however, continue to do so as the overlap
harnessing big data for decision- is laced with potholes. The biggest between different regulatory authorities
making mentioned by the survey impediment to effective decision- is rationalised. “Historically, you could
respondents are many and varied. making using big data, cited by 56% of say the islands of data provided some
survey respondents, is “organisational sort of job security,” says Mr Knorr
silos”. This appears especially the case of Citi. “If different areas have their
for large firms—those with annual own vernacular, then they keep to
Perceived benefits of revenue in excess of $10 billion—whose themselves and avoid transparency.
harnessing big data executives are more likely to cite silos as That has obviously broken down, mainly
for decision-making a problem (72%) than smaller firms with through the regulatory efforts to ensure
less than $500 million in revenue (43%). that the financial services industry can
“More complete have a consistent, end-to-end data
understanding of model that’s easily understood and
market conditions The intractable silos can relate the various transactions
and evolving and products across the board.”
business trends” The business process reengineering
(BPR) movement of the 1990s— Silos may also be eroded over time
“Better business led by Michael Hammer and by what Kurt Schlegel, a research vice
investment decisions” Thomas Davenport—attempted to president at Gartner, an analyst firm,
eradicate function silos. By mapping calls “trusted data aggregators”. He
“More accurate and processes (eg, “fulfil order”) that points to aggregators which collect
precise responses to ran “horizontally” through several data that different firms (often in
customer needs” functions (sales, distribution, accounts the same industry) can access and
receivable), duplicated tasks and other analyse for their own purposes. But
“Consistency of inefficiencies were identified and Mr Schlegel believes that the trusted
decision making eradicated, and data was made to flow data aggregator model can also work
and greater group more easily across function boundaries. within organisations themselves. And
participation in BPR was given a boost by the arrival even where data protection or privacy
shared decisions” of enterprise resource planning (ERP) laws prevent a given department
software which automated a number from revealing personal information,
“Focusing resources of common business processes. an aggregator could anonymise the
more efficiently for However, while BPR undoubtedly data and make it available to other
optimal returns” improved efficiency and made the inner departments.
machinations of functions visible—

56%
“Faster growth often for the first time—the “vertical”
of my business function silos were soon replaced by
(+20% per year)” “horizontal” application silos. Before,
data was trapped in functions; now
“Competitive it is trapped in ERP, CRM (customer
advantage (new data- relationship management) and SCM
driven services)” (supply chain management) systems.
of survey respondents cited
“Common basis—one To some extent, increasing
true starting point regulation, especially in the financial “organisational silos” are
for evaluation” services, pharmaceuticals and the biggest impediment to
telecommunications industries, has effective decision-making
“Better risk begun to erode data silos and will
using big data.
management”

18


Across all sectors, “organisational silos” are the biggest impediment to using big
data for effective decision-making

Survey Question: What are your organisation’s three biggest impediments to using big data for
effective decision-making?
[Select up to three options]

65.8%
63.0%
59.7%
57.1% 58.2%
55.7%
54.3% 54.5% 54.3%
52.6%
50.6% 50.0%
48.4%
44.3% 45.5%
43.7% 43.5%
40.0%
36.8% 37.1% 37.0%

Too many “silos”—data is not Shortage of skilled people to The time taken to analyse large
pooled for the benefit of the entire analyse the data properly. data sets.
organisation.

47.8% 48.4% 49.1%
45.7%
41.7% 41.8%
39.1%
36.8%
34.9% 34.3% 34.2%
32.9%

27.4%
24.3%

18.4% 20.0%
17.1%
14.5%
13.0% 13.0%
10.9%

Unstructured content in big data Big data is not viewed The high cost of storing and
is too difficult to interpret. sufficiently strategically manipulating large data sets.
by senior management.
41.3%

Total

Financial Sector
17.1%
14.7%
12.7% Energy & Natural Resources
10.9%
7.9% 8.1% 7.9% 8.1%
4.4% 4.3% 4.3%
1.8%
4.3%
Consumer goods & retail

IT & Technology

Manufacturing

Big data sets are too complex to Something not on this Healthcare & Pharmacy
collect and store. list (please specify).

19


Finding the right skills
The second big impediment to making and mathematics students, has been analytics”—where data sets are loaded
better decisions with big data is the running for 12 years in the US and is used into memory (RAM), making analysis
dearth of talented people to analyse in 18,000 schools; it will be offered to UK much faster—become more refined
it, mentioned by 51% of respondents. schools, for free, from March 2012. SAS and widely deployed, decision-making
For consumer goods and retail firms it has also developed advanced analytics at the operational and tactical level, at
is the single toughest obstacle, cited by courses with a number of universities, least, is likely also to become faster.
two-thirds of respondents from those including Centennial College, Canada,
sectors. North Carolina State University and
Saint Joseph’s University, Philadelphia,
“In terms of modelling, there is to provide the next generation of data
going to be a considerable shortage analysts.
[of specialists],” says Professor K
Sudhir, James L. Frank ‘32 professor
of marketing at Yale School of
Management. “As a nation we generally
The time factor
find math and sciences less exciting, The time it takes to analyse large
and I think people have been moving data sets is seen as another major
away from this to ‘softer’ sciences. impediment to more effective use of
Clearly, there is a shortfall, especially big data in decision-making. “I think
in the analyst domain, and it is going to big data is going to stimulate the
continue unless we systemically fix it.” need for more CPU [microprocessor]
Bill Ruh of GE agrees. “There is going to power, because people are going to
be a war for this kind of talent in the next get very creative and they’re going
five years,” he says. to invent new algorithms, and we’re
going to say ‘My God, everything’s
Aside from a master’s degree or PhD in slow again’,” says Mr Ruh of GE. “We
economics, mathematics, physics, or are going to have to redo our compute
other relevant field of science, analysts and storage architectures because they
are also expected to have in-depth will not work where all this is going.”
domain knowledge—something
which usually takes years to acquire. Most of the survey respondents have
Interviewees for this report also say not experienced a slowing of decision-
that the ideal analyst should have an making due to having to process
ability to communicate complex ideas large quantities of data. Only 7% say
in a simple manner and should be that it has slowed down decision-
customer-focused. Finding people with making significantly, while 35% say
all of these abilities is never going to be it has slowed it but only moderately.
easy, and retaining them is going to be (Respondents from transport,
even harder as the benefits of big data government, telecommunications
become apparent to more firms. and education suggest a greater
deceleration of decision-making than
Technology companies recognise the other sectors.) The impediment must
problem and are working with schools be, then, not that decision-making
and universities to develop these much is slowing, but that it is not getting
needed skills. For example, SAS, a faster. This seems to be borne out by
business analytics software firm based the fact that the vast majority (85%)
in Cary, North Carolina, developed of executives believe that the issue is
Curriculum Pathways, a web-based not the growing volumes of data, but
tool for teaching data analytics to high rather being able to analyse and act
school students. The course, aimed on data in real-time. As “in memory
at science, technology, engineering

20


85% of respondents say the issue is not about volume but the ability to analyse and
act on the data in real time

“The issue for us is now not the growing volumes of data, but rather being able to analyse and act
on data in real-time.”


100%

90%

80%

70%

60%

50%

40%

30%

20%

10%

Total Financial Energy & Consumer IT & Manufacturing Healthcare
Sector Resources Technology

Total Financial Energy & Consumer
Sector Resources
28.7% 21.7% 30.4% 37.8%
56.1% 62.3% 63.0% 48.6%
10.1% 10.1% 4.3% 8.1%
1.3% 1.4% 0.0% 0.0%
3.7% 4.3% 2.2% 5.4%

IT & Manufacturing Healthcare
Consumer
30.6% 36.4% 26.4%
53.2% 54.5% 62.2%
11.3% 7.3% 6.7%
3.2% 1.8% 2.2%
1.6% 0.0% 4.4%

21


Conclusion
Professor Alex Pentland, director of the heavy industry, especially in areas such
Human Dynamics Laboratory at MIT, as energy production and distribution
says big data is turning the process of (“smart grids”) and transportation
decision-making inside out3. Instead of (“smart cars”, etc), excessive automation
starting with a question or hypothesis, of business processes can hamper
people “data mine” to see what flexibility. Besides, the growing post-
patterns they can find. If the patterns financial-crisis regulation calling for
reveal a business opportunity or a greater accountability requires humans
threat, then a decision is made about to ultimately make the decisions.
how to act on the information. Prosecutors cannot put an algorithm in
the dock.
This is certainly true, but improvements The financial crisis has also led to calls
in computing power and artificial for greater transparency. As the survey
intelligence systems mean that asking shows, people are increasingly wary
direct questions of big data and getting of business decisions based purely
an answer, in real time, is now a reality on intuition and experience. Even if a
(see WellPoint case study). Although sizeable minority agree that business
these systems are still very costly and managers have a better feel for
not widely deployed, this research business decisions than analytics will
suggests that the appetite for real-time ever provide, managers will increasingly
decision-making is huge. And when need to show how they arrived at their
there is a business demand, it is only decision. And big data will provide a
a matter of time before the need if post-decision review—was it a good
fulfilled. decision or not? As one of the survey
participants puts it, using big data for
Most of the executives polled for this decision-making will lead to “better
report are also optimistic about the cost decisions; better consensus; better
reductions and efficiencies that can be execution”.
had from automating decision-making
using big data. While there is certainly
much scope for decision-automation in

23

About Capgemini
With around 120,000 people in 40 countries, Capgemini is one of the world’s foremost providers
of consulting, technology and outsourcing services. The Group reported 2011 global revenues of
EUR 9.7 billion.
Together with its clients, Capgemini creates and delivers business and technology solutions that
fit their needs and drive the results they want. A deeply multicultural organization, Capgemini
has developed its own way of working, the Collaborative Business Experience™, and draws on
Rightshore®, its worldwide delivery model.
Rightshore® is a trademark belonging to Capgemini

More information about our services, offices and research is available at

www.capgemini.com

Written by

The Deciding Factor: How big data is transforming business decision-making

Empfohlen

Empfohlen

Weitere ähnliche Inhalte

Was ist angesagt?

Was ist angesagt? (20)

Andere mochten auch

Andere mochten auch (20)

Ähnlich wie The Deciding Factor: How big data is transforming business decision-making

Ähnlich wie The Deciding Factor: How big data is transforming business decision-making (20)

Mehr von Capgemini

Mehr von Capgemini (20)

Kürzlich hochgeladen

Kürzlich hochgeladen (20)

The Deciding Factor: How big data is transforming business decision-making