Você está na página 1de 17

MAKING BUSINESS ANALYTICS WORK:

Lessons from Effective Analytics Users


Introduction

Business analytics is ascendant. Thats one of the most important conclusions to arise from
the SAS Power Series, which was held around the country during the summer and fall of 2011.
Tom Davenport, co-author of Competing on Analytics and arguably the nations foremost pro-
ponent of using business analytics for organizational decision making, attended these events
and captured the mood in a blog post, Analytics Are Everywhere.

Over the course of the conferences Davenport encountered a startling range of analytics ini-
tiatives being deployed across industries as diverse as retail, telecommunications, financial
services, publishing, advertising, government and health care. His conclusion: There is as
much variety in analytical programs and initiatives as there is in business and organizational
life in general.

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
2
The State of Analytics

Davenports observation that analytics are every- The Bloomberg Businessweek survey suggests that
where was also reflected in a survey of more than not only do such characteristics exist, but they can be
900 business executives worldwide, sponsored by described in very practical terms. As a result, we can
SAS and conducted by Bloomberg Businessweek identify 10 specific steps for making business analytics
Research Services during April and May 2011. For work better in organizations of any size, in any industry
instance, according to the survey, fully 97 percent of and at any level of analytical sophistication. For con-
companies with revenues of more than $100 million venience, we group these 10 steps into three major
said that they were using some form of business ana- categories:
lytics, up from 90 percent just two years ago.
Develop pro-analytics practices. The companies
But even as it validated the widespread reliance on that are most effective in their use of business
business analytics, the survey revealed that a steep analytics tend to be those that make the greatest
learning curve remains in using and benefiting from effort to actually employ business analytics in their
analytics. Among the key findings in this regard: Only organizational decision making.
one out of four organizations believes that its use of
business analytics has been very effective in helping Build a pro-analytics infrastructure. Those companies
to make decisions. As the first white paper in this se- that have been most successful in their use of
ries noted, this finding is a far cry from the competitive business analytics have built an operational in-
edge promised in all the hype around analytics. (The frastructure within their organizations that actively
full white paper can be viewed at sas.com/bbw1). supports the effective use of business analytics.

That white paper a detailed assessment of the cur- Create a pro-analytics culture. The most successful
rent state of business analytics concluded that the users of business analytics have the people in place
potential of business analytics was not overstated. who are committed to and capable of effectively employ-
However, rather than being the mature technology ing business analytics in their organizational decision
that it often is assumed to be, business analytics is making.
still in an emerging state. Organizations seeking
We conclude this white paper by examining a related
to make better use of business analytics stand to
and equally important issue: Is there an optimal balance
learn much from looking at the practices of organiza-
between intuition-based experience and data-based
tions that are already employing them successfully.
analytics in making organizational decisions? Before
we get to that question, however, we first review the
Making Analytics Work 10 practical steps for making analytics work better.

This white paper seeks to determine why some companies


are more effective users of business analytics than Develop Pro-Analytics Practices
others. In other words, are there certain behaviors or
Effective analytics users have developed pro-analytics
characteristics that consistently lead to greater suc-
practices, most readily apparent in four areas: (1) an
cess in the use of analytics?
overall reliance on business analytics; (2) the integra-

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
3
tion of business analytics across the organization; (3) the most effective users of analytics. For example,
the use of business analytics to address specific busi- 86 percent of respondents that rate themselves very
ness tasks or challenges; and (4) the use of specific effective analytics users say that their organizations
analytical tools. rely heavily on data and metrics to make business
decisions, compared with only 45 percent of those who
We have identified four concrete steps for developing deem themselves less effective. (Figure 2) Not surpris-
pro-analytics practices: ingly, almost all (97 percent) of the very effective
Step 1. Expand the use of business analytics companies that rely most heavily on business analytics
where feasible. The Bloomberg Businessweek sur- characterize themselves in this way. However, even
vey shows that companies that rely more on business among very effective companies that are much less
analytics in making decisions are significantly more reliant on business analytics, 71 percent say that they
likely to say that their use of business analytics has rely heavily on data and metrics to make business
been very effective in helping to make those deci- decisions.
sions. Overall, high-reliance organizations are more Figure 2. Heavy Reliance on Analytics
than twice as likely as low-reliance companies (38
percent vs. 16 percent) to say that business analytics 100%
has been very effective in assisting their business 90% 86%
decision making. Conversely, low-reliance firms are 80%
three times more likely than high-reliance firms (13 70%
percent vs. 4 percent) to say that business analytics 60%
has been not effective at all. (Figure 1) 50% 45%
Figure 1. Relationship Between Reliance and 40%
Effectiveness 30%
38% 20%
40%
10%
35%
0%
30% Very Effective Users Less-Effective Users

25%
Very Effective Less-Effective Users
20% 16%
15% 13%
The lesson for companies that wish to be more effective
10% analytics users is straightforward: Shift analytics use
4% from occasional to routine. With day-to-day practice
5%
and reliance comes greater effectiveness.
0%
High-Reliance Low-Reliance Step 2. Integrate across the organization. Business
analytics has been integrated across the organization
Very Effective Not Effective
in 57 percent of companies that rate themselves very
effective in their use of analytics. By contrast, among
Indeed, reliance on business analytics tends to companies that have been less effective in their use
become second nature in the companies that are

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
4
of business analytics, less than half this proportion in their use of business analytics as compared with
just 28 percent of firms have integrated busi- less-effective analytics users. (Figure 4) For instance:
ness analytics organization wide. Looked at from the
Fully 78 percent of companies that are very
opposite perspective, isolation of business analytics to
effective in their use of business analytics employ
a single department or function commonly referred
business analytics in strategy and planning, as
to as siloing is more than three times as common
compared with just 55 percent of other firms.
(31 percent vs. 9 percent) among these less-effective
users of analytics than it is in companies that are very Some 69 percent of companies that are very
effective in their analytics usage. (Figure 3) effective in their use of business analytics employ
Figure 3. Integration Across the Organization analytics in finance, versus 49 percent of other firms.

60% 57% Some 68 percent of very effective companies


use analytics in marketing, versus 48 percent of
50%
other companies.
40%
28%
31% Sixty-four percent of very effective firms employ
30%
analytics in sales, as compared with just 44 per-
20% cent of other firms.
9%
10%
Sixty percent of very effective companies use
0% analytics in operations and supply chain manage-
Very Effective Users Other Companies
ment, versus 41 percent of other firms.

Integrated Isolated Fifty-seven percent of very effective firms employ


analytics in product development, as compared
Siloing of data and functionality is hardly restricted with just 34 percent of other firms.
to business analytics, as the condition affects a wide
Figure 4. Use of Business Analytics Functions
array of corporate information and operations, from
resource planning to customer-experience manage- Very Effective Other
Users Companies
ment. In these circumstances, siloing often becomes
Strategy/Planning 78% 55%
a persistent and challenging problem to solve. What Finance 69% 49%
is required more than anything is an organizational Marketing 68% 48%
commitment to break down silos, employing such Sales 64% 44%
Information Technology/Management 64% 48%
strategies as the creation of cross-divisional data Operations/Supply Chain Management 60% 41%
teams or organizational centers of excellence to help Product Development 57% 34%
Customer Service/Support 54% 37%
share data and processes across the organization. Human Resources 40% 29%

Step 3. Bring analytics to task. The use of business Business analytics systems, like the data that underlies
analytics tends (by an average of about 20 percentage them, can be complex, and it can be difficult for orga-
points) to be more highly task- or mission-focused in nizations to master a wide range of analytical functions
those companies that rate themselves very effective right away, even in cases in which the data is available.

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
5
So the best choice for organizations that are currently Forty-three percent of very effective firms
less effective in their use of business analytics may be employ optimization tools, versus 24 percent
to begin with just one or two business analytics initia- of other companies.
tives and build from there.
And 34 percent of very effective companies employ
Jayson Tipp, Vice President of Analytics and CRM at Web analytics, versus just 23 percent of other firms.
DVD distributor Redbox, calls this step-by-step pro-
cess offering analytical saltines to an organization. A Figure 5. Use of Business Analytics Tools
full customer data warehouse and analytical infrastruc- Very Effective Other
Users Companies
ture may not yet be available, he says, but there are
Spreadsheets 74% 59%
smaller analyses that can be undertaken in the short Business reporting/KPIs/dashboards 72% 52%
term using available data sets. To someone stranded Forecasting 71% 50%
Query and Analysis 57% 35%
on a desert island, Tipp explains, just some saltines General statistics 53% 37%
can taste like steak. We want to eventually provide Data/Text Mining 50% 36%
Simulations and scenario development 46% 27%
the steak, but for now we can at least make some Model management 43% 24%
Optimization 43% 24%
saltines available. Web analytics 34% 23%
Interactive data visualization 31% 15%
Step 4. Use the tools. In addition to deploying business Social media analytics 22% 21%
Text/audio/video analytics 16% 12%
analytics to address specific business tasks and chal- None of the above 1% 2%
lenges, companies ranking themselves very effective
in their use of business analytics also tend (again by The message for organizations that are currently less
an average of about 20 percentage points) to be more effective in their use of business analytics is plain: Be
aggressive in learning and using specific analytical bolder. Because simple tools like spreadsheets are
tools. (Figure 5) For instance: so familiar to analysts and executives alike, many
companies remain content with these relatively un-
Fully 72 percent of companies that are very sophisticated analytical tools, and they fail to extend
effective in their use of business analytics use their capabilities. But once the decision to branch out
business reporting, key performance indicators is made, the results can be truly impressive.
and dashboards, versus just 52 percent of less
effective analytics users.
Build a Pro-Analytics Infrastructure
Some 71 percent of very effective companies
Organizations that have been most successful in
employ forecasting tools, as compared with just
using business analytics have built an infrastructure
50 percent of other companies.
that matches their pro-analytics practices. Typically
Some 50 percent of very effective firms undertake an analytics infrastructure incorporates three key ele-
data and text mining, versus 36 percent of other ments: (1) access to data; (2) access to technology;
companies. and (3) defined data-management processes.

Forty-six percent of very effective companies Step 5. Create a data-management strategy that
use simulations and scenario development, as includes ready access to data. Studies have repeat-
compared with 27 percent of other companies. edly shown and the current survey validates that
the lack of data quality, integrity and consistency is

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
6
one of the main impediments to the adoption of busi- These findings make sense. If relevant data is made
ness analytics within organizations. For instance, in widely available as part of a comprehensive data-
the Bloomberg Businessweek survey, fully 34 percent management strategy, it can more easily be used as
of companies that were very effective in their use of an input into a business analytics system. On the other
business analytics said that the quality, integrity and hand, if data is restricted, fragmented or otherwise not
consistency of data was the most difficult business available, if will be difficult even for those organizations
analytics challenge they faced. Respondents from that are inclined to use business analytics actually to
other companies also reported data quality, integrity do so. The message for organizations that are cur-
and consistency problems as their most vexing chal- rently less effective in their use of business analytics
lenges in implementing business analytics, although should be clear: Make sure that the data necessary
for performing analytics is readily available to all who
the percentage naming this issue was somewhat lower
might need it.
(28 percent) than for very effective analytics users.

While such data-related problems can make it more Figure 6. Access to Data
difficult for organizations to employ business analyt-
80% 75%
72%
ics in their decision making, having greater access to
70%
data has an unquestionably positive influence on the 61%
60%
effectiveness of business analytics efforts. (Figure 6)
For instance: 50%
40%
37%
40% 34%
Seventy-five percent of respondents that identify
30%
themselves as very effective users of business
20%
analytics (i.e., very effective companies) say
10%
that business information within their organiza-
tions is readily available to those who need it, as 0%
Info Readily Central Data Info Shared
compared to 37 percent of respondents that deem Available Source Throughout
themselves less effective in their use of analytics
(i.e., less effective companies).
Very Effective Users Other Companies

Almost twice as many respondents in very effective


companies as in less-effective companies (61 Step 6. Deploy the necessary technology. By its
percent vs. 34 percent) say that people in their nature, business analytics is a very data-intensive
organizations have access to integrated and high- activity, and properly analyzing and deploying this data
quality information across all business units and inevitably requires technology often, very sophisti-
functions from a central source, such as a data cated technology. Of course, much of what is described
warehouse. as business analytics is fairly low-tech. For instance,
in the Bloomberg Businessweek survey, respondents
And almost twice as many respondents in very
from 63 percent of very effective companies and from
effective companies as in less-effective compa-
54 percent of less-effective companies said that their
nies (72 percent vs. 40 percent) say that data and
firms considered standard electronic spreadsheets to
information are shared across their organizations.

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
7
be a business analytics tool. Surprisingly, respondents ily available within an organization, it can more readily
from 75 percent of the very effective companies that be employed to support business analytics activities.
are the most intensive users of business analytics By contrast, if the necessary technology is not present,
said that their companies regard spreadsheets as an it will be difficult even for those organizations that are
analytics tool. committed to using business analytics to make use of
it in their business decision making. Again, the lesson
Nevertheless, any company with even a modicum of for organizations that have been less effective in their
experience in the business analytics field surely un- business-analytics efforts is plain: Make the necessary
derstands that the most informative business analytics technology available to conduct the kind of business
efforts require the use of computer software and other analytics that would be most helpful to meeting the
technology that is far more complex and capable than organizations goals and needs.
standard-issue spreadsheets. Companies that rate
themselves very effective in using business analyt- Figure 7. Business Analytics Technology
ics, in particular, appear to recognize this fact, as most
80%
have gone to great lengths to make more sophisticated 71% 71%
70% 63%
technology widely available within their organizations.
(Figure 7) For instance: 60%
50%
More than twice as many respondents in very 38%
40% 33% 34%
effective companies as in less-effective compa-
30%
nies (71 percent vs. 33 percent) say that the use
20%
of advanced analytics software, such as software
10%
for data mining, statistical forecasting and predic-
tive analysis, is integrated into relevant business 0%
Advanced Technology Standardized
processes within their organizations. Software Used Available Technology

Nearly twice as many respondents in very effective


companies as in less-effective companies (71 per- Very Effective Users Other Companies

cent vs. 38 percent) say that employees who need


access to business information have the technol- Step 7. Develop formal data-management processes.
ogy and processes available to them to retrieve As is the case with many corporate initiatives, the
that business information in a timely matter. administrative environment often can make or break an
organizations efforts to rely more heavily on business
Almost twice as many respondents in very analytics in their organizational decision making. As a
effective companies as in less-effective firms (63 consequence, ensuring that appropriate data-manage-
percent vs. 34 percent) say that their organizations ment processes are in place can be a crucial step. The
have standardized technology for accessing, inte- Bloomberg Businessweek survey bears out this point.
grating and analyzing information from all business In general, the companies that are most effective in
functions. their use of business analytics have put in place three
key sets of data-management processes: (1) general
As with the availability of the relevant data, these findings
data-governance rules and policies; (2) defined data
also make sense. If the necessary technology is read-

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
8
stewardship; and (3) master data definitions. (Figure 8) tion to implement and enforce them. They also require
Specifically: vigilance in keeping the rules, policies and definitions
up to date. Companies that are currently less effective
About half again as many respondents in very in their use of business analytics therefore can take a
effective companies as in less-effective com- major step forward by solidifying the data-management
panies (64 percent vs. 41 percent) say that their policies that govern the use of analytics within their
organizations have a formal set of rules and poli- organizations.
cies in place to govern information-management
processes across the organization.
Create a Pro-Analytics Culture
Almost twice as many respondents in very effective
There remains one final element of success in
companies as in less-effective companies (65 per-
employing business analytics. Companies that are
cent to 36 percent) say that their organizations
most effective in using analytics to support their or-
have business people assigned as data stewards
ganizational decision making do more than develop
to ensure that the organization has a single set of
pro-analytics practices, and they do more than put the
accurate, high-quality data that is used for deci-
right infrastructure in place. They also build a strongly
sion making.
pro-analytics organizational culture that supports and
And twice as many respondents in very effective encourages the use of business analytics. As the first
companies as in less-effective companies (60 white paper in this series emphasized, Culture plays a
percent to 31 percent) say that their organizations critical role in the effective use of business analytics.
have a standard set of master data definitions that This culture can be seen in three key areas: (1) senior
everyone follows. executive support; (2) day-to-day reliance; and (3) the
requisite analytical talent.
Figure 8. Data-Management Processes
Step 8. Secure senior executive buy-in. Respondents
70% 64% 65% from approximately half (49 percent) of companies that
60%
60% identify themselves as very effective in their use of
50% business analytics say that, in their organizations, the
41%
36% use of business analytics has been primarily a grass-
40%
31% roots effort. However, regardless of where the initiative
30%
originates, high-level executive support appears to be
20%
very helpful in promoting the long-term success of busi-
10% ness analytics efforts. (Figure 9) All told, nearly twice
0% as many respondents in very effective companies as
Formal Set of Data Stewards Master Data in less-effective companies (86 percent vs. 49 percent)
Rules in Place Assigned Definitions
say that executive managers in their organization place
a great deal of trust in the results from business analyt-
Very Effective Users Other Companies
ics. This trust consequently translates into executive
support of business analytics, as well as the provision
Data-management policies do not have to be complex, of the resources, budget and talent to implement busi-
but they do require a commitment within the organiza- ness analytics on an ongoing basis.

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
9
Of course, securing executive buy-in is not always Two implications arise from these findings for less-
easy, and so bottom-up initiatives will continue to play effective companies that want business analytics to play
a major role in the successful implementation of busi- a greater role in their organizations decision making. One
ness analytics efforts. Still, it remains worthwhile pur- is to work hard to deliver positive results from analytics,
suing senior executive support, even if such support even if at first they are small. Second is to communicate
takes time to fully materialize. For the lesson from the those positive results throughout the organization to build
survey is clear: The more enthusiastic and committed a positive perception of business analytics.
the organizations senior executives are to business
analytics initiatives, the more successful those initia- Figure 10. Positive Impact from Analytics
tives are likely to be. 100% 90%
90%
Figure 9. Senior Executive Trust of Analytics
80%
100% 70%
90% 86% 60% 54%
80% 50%
70% 40%
60% 30%
49%
50% 20%
40% 10%
30% 0%
20% Very Effective Users Other Companies
10%
0% Very Effective Users Other Companies
Very Effective Users Other Companies
Step 10. Hire and develop the right analytical talent.
Very Effective Users Other Companies Data, technology, processes and executive support
are critical aspects of the analytics environment, but
Step 9. Deliver and communicate value. One of these elements are not very useful on their own with-
the most profound cultural differences among very out another critical component: people. For without
effective and less-effective companies in terms skilled and committed analytical talent, neither the data
of business analytics lies in the perception of value nor the technology will be put to optimal use. Profes-
that analytics provides. (Figure 10) For instance, 90 sionals who are committed to and skilled in business
percent of respondents from very effective compa- analytics can transform raw and context-less data into
nies strongly agree with the statement that the use of definitive, useful information. As a result, the presence
business analytics has had a positive impact on the of the right analytical talent is one of the hallmarks of
way their organizations do business. Interestingly, the organizations that are very effective in their use of
results vary little according to whether the company is business analytics. (Figure 11) For instance:
highly reliant on analytics (96 percent reporting a posi-
Nearly twice as many respondents in very effective
tive impact) or relies little on analytics (84 percent re-
companies as in less-effective companies (71
porting a positive impact). By contrast, just 54 percent
percent vs. 43 percent) say that their organiza-
of respondents from less-effective companies say that
tions place a premium on having workers with
they have seen a positive impact from analytics.
strong analytical capabilities.

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
10
This orientation translates into action: Respondents ing basis to undertake the analytical number-crunching
in very effective companies also are almost that both informs the organization and makes it possible
twice as likely (34 percent vs. 19 percent) as are to reap significant benefits from business analytics.
those in less-effective companies to say that their Securing that analytical talent involves recruiting and
organizations have the right analytical talent in hiring the right talent, then training these individuals in
the skills and processes required for successfully im-
place to implement business analytics.
plementing business analytics within the organization.
Respondents in less-effective companies are five
times as likely as those in very effective compa- Find the Right Balance
nies (15 percent vs. 3 percent) to say that their Both data and judgment obviously play a key role in
organizations do not have a handle on the ana- most major business decisions. But what is the right
lytical talent necessary to use business analytics balance? As the first paper in this series stated, Where
effectively. the optimal balance lies depends on a number of fac-
tors, including the organization, the tools, the people
Figure 11. Business Analytics Talent and, of course, the decision itself.
80%
71% In other words, the optimal balance between data and
70% experience (or judgment) is situation- and context-
60% dependent. But do the findings from the Bloomberg
50% 43% Businessweek survey indicate an ideal balance in the
40% 34% companies that are the most effective users of business
analytics? As you might expect, companies that deem
30%
19% themselves very effective in their use of business
20% 15%
analytics tend to place greater reliance on business
10% 3% analytics in their decision making than do other com-
0% panies. Moreover, within the subset of very effective
Premium on Right Talent Dont Know companies, those firms that are the most highly reliant
Analytic Talent in Place Talent Needed
on analytics place the majority of the weight in decision
making on business analytics. Specifically:
Very Effective Users Other Companies
The balance between experience and analytics
On a related note, Tom Davenport emphasizes the differs substantially among the two groups of
importance of analytical talent in his report on the first companies. Very effective companies tend to
SAS Power Series meeting. He says that he came rely almost equally on experience and business
away from that gathering with one big idea: the power analytics in their decision making, with a mix of
of the analytics catalyst. Says Davenport: If you want 53 percent experience and 47 percent analytics.
to get analytics and fact-based decision making going Among less-effective companies, the balance is
in your organization, hire an experienced analytical 62 percent experience and 38 percent business
executive to catalyze your efforts. These are people analytics. While not a profound shift, the differ-
who know business, know analytics and know how ences between the two groups of companies
to get things moving in organizations through trusted
are nevertheless significant, and indicate that ef-
relationships and rapid, reliable delivery.
fectiveness in using business analytics is clearly
Experienced analysts are necessary to get the process associated with greater reliance on analytics in
started, and skilled analysts are needed on an ongo- decision making. (Figure 12)

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
11
Figure 12. Balance of Analytics vs. Experience we never make decisions based on analytics alone.
The first step, he says, is to employ some combination
70%
62% of intuition and business experience in order to pro-
60% 53% spectively gauge how much time to spend on applying
50% 47% business analytics techniques to the flow of data a
38% judgment based, in part, on a comparison of the costs
40%
30% of analysis versus the costs of making an incorrect
decision. Then, once the appropriate level of effort
20%
is defined, business analytics is deployed to answer
10%
the relevant business questions to the extent that the
0% analytical framework allows.
Very Effective Users Other Companies
Figure 13. Relationship Between Analytics
Very Effective Users Other Companies
Reliance and Analytics Balance
Further, among very effective companies, 80% 73%
there is no fixed balance between the reliance 69%
70%
on intuition and experience on the one hand and
60%
on quantitative data and analysis on the other.
50%
Predictably, in those companies that rely most
40%
heavily on business analytics, decision making 31%
30% 27%
primarily depends on quantitative analysis and
data, with the average decision-making balance 20%
a mix of 31 percent experience and 69 percent 10%
business analytics. In contrast, in companies that 0%
rely less heavily on business analytics, the aver- High-Reliance Low-Reliance
age decision-making mix is nearly the opposite:
73 percent experience and 27 percent business Analytics Experience
analytics. (Figure 13)
Moreover, as Tipps comments suggest, more is not
What is the source of this disparity? Tipp views the always better in the analytical realm. While business
discrepancy as more artifact than intention, and sug- analytics is being used with increasing frequency,
gests that analytical inputs should be viewed in terms the Bloomberg Businessweek survey indicates that it
of a process rather than a numerical balance (see should be used judiciously as well. Even among the
sidebar). At Redbox, he notes, there are no fixed very effective users of business analytics, respon-
guidelines for how much data and how much intuition dents from some 27 percent of companies say that
and experience to inject into a given business deci- their firms previously experienced negative outcomes
sion. Instead, the optimal use of business analytics from decision making informed by business analytics,
within Redbox has relied more on the appropriate and respondents from 33 percent of these very ef-
sequencing of decision-making inputs than it has on fective firms said that their use of business analytics
any normative balance among them. had contributed to paralysis by analysis within their
organizations. Although these adverse consequences
Its true that business analytics is almost always the resulted in only a minority of cases, they nevertheless
backbone of our decision making, Tipp explains, but

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
12
suggest that finding the right approach to using business that we can set forth a set of 10 concrete and practical
analytics involves knowing when not to rely on analytics. steps that organizations wishing to improve their effec-
tiveness in using business analytics can take.
Practical Steps for Promoting
Business Analytics While none of these steps is necessarily easy, all can
be accomplished. And those organizations that commit
Business analytics is a framework with great potential themselves to implementing them will find that, before
for improving organizational decision making in a wide long, greater success in business analytics will be well
variety of functional areas. And while the Bloomberg within their reach and that they, too, will be able to join
Businessweek survey indicates that business analyt-
the ranks of their fellow organizations already proving
ics is still in its emerging stages, enough is known
to be very effective in their use of business analytics.

10 Practical Steps for Making Analytics Work


1. Expand the use of business 6. Deploy the necessary analytical
analytics where feasible. technology.

2. Integrate analytics across the 7. Develop formal data-management


organization. processes.

3. Deploy analytics on specific 8. Secure executive support.


business tasks.
9. Deliver and communicate value.
4. Use a variety of analytics tools,
including more sophisticated ones. 10. Hire and develop the right
analytical talent.
5. Create a data-management strategy
that includes ready access to data.

Making Business Analytics Work:


Lessons from Effective Analytics Users
Prepared by Bloomberg Businessweek Research Services
13
Addendum:
Redbox: An Analytics in Action Case Study

Vending machines are nothing new to American consumers theyve been around for more than 100 years, dispensing
everything from snacks to sandwiches to soda pop. In an age where DVD movies have become Americans primary form of
home entertainment, the idea of automated kiosks that supply popular DVD rentals seems like such a natural extension of the
vending-machine concept that one might wonder why the automated kiosks didnt appear sooner.

In fact, they did but in a much different form. Redbox began life nearly a decade ago as a rather prosaic premise: a conve-
nience store in a box. The idea was to make available the staples of the typical convenience store in a mere 150-square-foot
space, allowing consumers to purchase frequently needed items without having to drive to the store itself.

Finding the Golden Nugget

Unfortunately, the convenience store in a box concept turned out not to be as popular with consumers as the original Redbox
team had hoped, and sales fell short of expectations. But one product within the vending machines portfolio stood out: DVD
rentals. In fact, the idea of automated DVD rentals proved to be so compelling and lucrative that Redbox executives soon
decided to winnow the original convenience store concept down to that single product line. In the process, the modern-day
Redbox kiosk was born.

In the half-dozen years since, the business has thrived. There were a mere handful of Redbox kiosks in 2002; now, there are
more than 33,000. Rentals are brisk, and its not uncommon even at a time when the traditional video-rental stores have
all but disappeared and online rental services are booming to see a dozen or more patrons lined up at the local Redbox
kiosk on a Friday night.

How did the Redbox team alight on this eminently successful formula? There was a bit of inspiration and inventiveness, no
doubt. But at the heart of the companys evolution was the careful analysis of business data, employing the techniques of a
sophisticated business process called business analytics. By relying on the business analytics toolbox as they studied the
results of the early Redbox trials, Redbox executives were able to do far more than inspect the companys bottom line. They
could examine in great detail the components of and the interrelationships among the companys costs, revenues and
profits. And they could efficiently explore a number of potential what-if scenarios that told them not only what was happening
now, but what was likely to take place in the future.

What became immediately clear from this analysis was that DVD rentals not only produced a high current margin combining
a relatively high return with relatively low maintenance costs but also showed promise for rapidly accelerating revenue
growth in a way that the early machines more commoditized offerings did not.

From Data to Dollars

Redboxs revenues seemed ready to explode, but capitalizing on this potential would prove to be anything but simple. Any
business faces operational challenges, of course, and thats especially true of startup ventures entering into a high-growth
phase. Redbox had some unique challenges to overcome, though. Kiosk maintenance and inventory changes were major
issues: In order to be profitable, the company needed to minimize the high labor costs associated with frequent kiosk service.

Addendum:
Redbox: An Analytics in Action Case Study
Prepared by Bloomberg Businessweek Research Services
14
Redbox also had to preserve strong relationships with its retailers so that the retailers themselves would take the initiative to
ensure that the kiosks were kept clean and free of obstructions and damage.

And then, most critically, there was the customer side of the equation. The Redbox team knew that convenience was a prime
selling point to the companys customers, and that any shortcomings in this area could thwart customer usage. As a result, the
Redbox operating system needed to seamlessly track DVD rentals, process payments and allow customers to return DVDs to
any Redbox unit all the while supplying business executives with a steady stream of insightful metrics like average rental
time, repeat usage and product-popularity curves. Moreover, these metrics needed to be analyzed on a region-by-region and,
sometimes, even on a machine-by-machine basis. Because of the frequent changes in consumer tastes, the data needed to
be provided in close to real time. These requirements, which were similar to those facing traditional video-rental stores, were
greatly accentuated for Redbox. Because each kiosk could accommodate only a very limited number of DVDs and DVD titles,
it was imperative for the company to maximize the revenue from each title and each individual machine.

To produce the data needed to support such decisions, Redbox constructed a sophisticated business analytics operation that
tracked not only aggregates, but also microdata, such as daily popularity indices, regional preferences and popularity-falloff
curves. According to Jayson Tipp, the head of Redboxs business analytics marketing unit, the ability to produce and assess
this type of operational and financial data in close to real time has long been one of the companys core competencies and
one of its major reasons for success. Such data, he says, has informed a great deal of both our strategic and day-to-day
decision making.

This deepening foray into business analytics been so beneficial that over the last year, Redbox has extended business
analytics to its marketing operations. Now, carefully honed data and analysis supports decisions in customer relationship
management, database marketing and marketing performance management, among others. As a result, Tipp says, We now
have a really strong and rigorous analytical framework for understanding how to drive marketing promotions that create a
great experience for consumers and drive the business forward in the ways that we want. At the same time, he notes that
business analytics has allowed the team to rule out a number of marketing options with lesser potential before significant
financial resources were committed to them.

Relying on Analytical Talent

One of the cornerstones of Redboxs success with business analytics has been the companys commitment to hiring and
developing the kind of analytical talent necessary to produce and act upon this highly informative data. According to Tipp,
there are a large number of really smart people at Redbox who have been highly focused on business analytics. Much of
the companys achievements in this arena have been due to the thought leadership provided by Gregg Kaplan, the original
Director of New Ventures at McDonalds and, later, CEO of Redbox (and now COO and President of CoinStar, the current
owner of Redbox).

Kaplan, Tipp says, is a very sharp thinker with a deeply analytical orientation. From the beginning, he insisted that the Redbox
executive team be very clear about the companys objectives and then find ways of measuring progress toward achieving
them. Originally, the team needed to review only a few electronic spreadsheets. But as transaction volume grew, more power-
ful tools were needed to extract the required information from the data, and the analytical environment became increasingly
complex. Making the investment in the right tools, Tipp says, helped the company to keep pace with these analytical needs.

Finding the Right Balance

Finding the right analytical balance was also essential, Tipp adds. There were no fixed guidelines for how much data and how
much intuition and experience to inject into a given business decision. At no level of the organization, he says, is there a
set rule as to where we leverage intuition and where we dont, though he notes, there are areas of the business that lend

Addendum:
Redbox: An Analytics in Action Case Study
Prepared by Bloomberg Businessweek Research Services
15
themselves to greater quantitative analysis, while there are other areas that are more intuition-driven. Regardless of the
specific balance, Tipp says, the key driving force is that people throughout the Redbox organization want to be as informed as
possible about the implications and timing of the decisions that they are called upon to make.

As a result of this perspective, the optimal use of business analytics within Redbox has relied more on the appropriate
sequencing of decision-making inputs than it has on any normative balance among them. Its true that business analytics is
almost always the backbone of our decision making, Tipp explains, but we never make decisions based on analytics alone.
The first step, he says, is to employ some combination of intuition and business experience in order to prospectively gauge
how much time to spend on applying business analytics techniques to the flow of data a judgment based, in part, on a
comparison of the costs of analysis versus the costs of making an incorrect decision. Once the appropriate level of effort is
defined, the heavy lifting of business analytics is deployed to answer the relevant business questions to the extent that the
analytical framework allows.

Its a process that requires finesse and experience to implement efficiently. But in the case of Redbox its one that has been
instrumental to the companys rapid and ongoing success.

Addendum:
Redbox: An Analytics in Action Case Study
Prepared by Bloomberg Businessweek Research Services
16
Copyright and Disclaimer Notices

Neither Bloomberg Businessweek, Bloomberg nor SAS makes any guarantees or warranties as to the accuracy or completeness of this
report. Neither Bloomberg Businessweek, Bloomberg or SAS shall be liable to the user or anyone else for any inaccuracy, error or omission,
regardless of cause, or for any damages resulting therefrom. In no event will Bloomberg Businessweek, Bloomberg, SAS, or any of their
third-party licensors be liable for any indirect, special or consequential damages, including but not limited to lost time, lost money, lost profits or
lost good will, whether in contract, tort, strict liability or otherwise, and whether or not such damages are foreseen or unforeseen with respect
to any use of this document. This document, or any portion thereof, may not be reproduced, transmitted, introduced into a retrieval system or
distributed without the written consent of Bloomberg and SAS.

Copyright 2012 Bloomberg and SAS. All rights reserved.

The names of actual companies and products mentioned herein may be the trademarks of their respective owners.

ELECTRONIC VERSION AVAILABLE


To see or use an electronic copy of this document in PDF format, please visit: sas.com/bbw2

Você também pode gostar