Introduction

It’s not what you look at that matters, it’s what

you see.
—Henry David Thoreau

P
rofessional writers and speakers like me live interesting lives. I’d hazard to
guess that most of us work from home, although some maintain proper
offices. And when you work from home, strange things can happen. For
one, it can become difficult to separate work from leisure. There’s no boss
looking over your shoulder to see if you’ve completed that TPS report. Did you
get that memo? If you want to take a nap in the early afternoon as I routinely
do, no one’s stopping you. In a way, people like me are always at work, even
though we’re not always working. It’s fair to say that the notion of work-life
balance can be challenging. Lines usually blur. Maybe they’re even obliterated.
In many ways, working from home could not be more different from
working for “the man.” Even today, many rigid corporate environments block
employees from visiting certain websites via services like Websense. And forget
the obvious sites (read: porn). At many companies, there’s no guarantee that
employees can access websites that serve legitimate business purposes, at
least without a call to the IT help desk to unblock them. Examples include
Twitter, Facebook, Tumblr, and Pinterest. Of course, many employees in indus­
trialized countries sport smartphones these days, minimizing the effectiveness
of the Websenses of the world. As a result, many companies have reluctantly
embraced the Bring Your Own Device movement. That genie is out of the
bottle.
We home-based employees, though, don’t have to worry about these types
of restrictions. No one stops us from wasting as much time as we want on
the Web, the golf course, or anywhere else for that matter. In an increasingly

cintro 3 January 28, 2014 9:50 PM

 4 ▸   B o o k Overview and Background

blurry world, though, what does it really mean to waste time? That’s a bit
existential. Let me rephrase: Are my tweets generally work related? Have they
changed over time? If so, how?

Adventures in Twitter Data Discovery

Twitter tells me that, since 2010, I have tweeted more than 17,000 times as of
this writing, or about ten times per day. I’d wager that more than 70 percent
of my tweets were work related. (Yes, I have been paid to tweet. Lamentably,
I don’t command Kim Kardashian-type rates for my 140 characters.* Maybe some
day.) Twitter has let me connect with interesting people and organizations,
many of whom you’ll meet in this book. In the course of researching this book,
I searched Twitter for a random sample of thoughts, typically with the hashtag
#dataviz. At least to me, Twitter is an exceptionally valuable business service
that I would gladly pay to use. While we’re at it, let’s put Twitter client Hoot-
Suite in that same boat.
At the same time, though, I unabashedly use Twitter for reasons that have
absolutely no connection to work. If you go to @philsimon and follow me
(please do), there’s a good chance that you’ll see a few tweets with #Rush and
#BreakingBad, my favorite band and TV show, respectively. What’s more,
I’ve tweeted many of these things during times and days when I probably
should have been working. I could delude myself, but I won’t. A few of my
favorite celebrities and athletes have engaged with me on Twitter, bringing a
smile to my face. I’ll say it: Twitter is fun.
But let’s stick with work here. Based on what I’m doing, I suspect that my
tweets have evolved over time, but how? It’s presumptuous to assume that the
content of my tweets is static. (I like to think that I have a dynamic personality.)
To answer this question, I could have accessed my archived tweets
via Twitter.com. The company made user data available for download in
December 2012. I could have thrown that data into Microsoft Excel or Access
and started manually looking for patterns. Knowing me, I would have created
a pivot table in Excel along with a pie chart or a basic bar graph. (Yes, I am a
geek and I always have been.) The entire process would have been pretty time
consuming even though I’ve been working with these productivity staples for
a long time. Let’s say that Twitter existed in 1998. If I wanted to visualize and
understand my tweets back then, I would have had to go the Microsoft route.
Of course, it’s not 1998 anymore. Answering these simple questions now
requires less thought and data analysis than you might expect. Technology
today is far more powerful, open, user-friendly, ubiquitous, and inexpensive
compared to the mid-1990s.

* Reportedly, a mind-blowing $10,000 per tweet.

cintro 4 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 5

Like many companies today, Twitter relies upon a relatively open applica-
tion programming interface (API).* At a high level, APIs allow devices, apps,
and Web services to easily interact with one another. They also facilitate the
near-instant flow of data. Lately, APIs have become all the rage. Myriad people
use them every day, whether they know it or not. Facebook, LinkedIn, Four-
Square, Google, and scores of other companies effectively use APIs for all sorts
of reasons. And forget massive tech companies with billion-dollar valuations.
Many start-ups are based on “the Twitter fire hose,” including the aforemen-
tioned HootSuite. Open APIs encourage development of third-party products
and services, a topic I discussed in great detail in The Age of the Platform.
One such service is Vizify, a start-up founded in 2011 and based in Port-
land, Oregon. The company is a proud graduate of both Seattle TechStars
and the Portland Seed Fund. I fittingly “met” company cofounder and CEO
Todd Silverstein over Twitter in June 2013 while researching this book. Vizify
quickly and easily lets users connect to different social networks like Facebook,
Twitter, FourSquare, and LinkedIn.
It took about three minutes for Vizify to pull my photos, education
history, current occupation, work history, home page, tweets, and other key
profile data that I’ve chosen to make publicly available. Of course, users aren’t
obligated to connect to any individual network. (I passed on FourSquare.)
After the initial load, users can easily remove pictures or other information
they would prefer not to share. By accessing open APIs, Vizify allows users to
create free and interactive visual profiles. Mine is shown in Figure I.1.
If you want to see my full multipage profile, go to https://www.vizify.com/
phil-simon. In case you’re wondering, users can change the colors on their
profiles. I went with that particular shade of green as a homage to Breaking Bad.
A snazzy visual profile is all fine and dandy, but it still didn’t answer my
Twitter question. Fortunately, Vizify also allowed me to effortlessly see the
evolution of my tweets over time. A screenshot from that part of my profile is
shown in Figure I.2.
Figure I.2 proved what I had suspected. First, I use Twitter for both business
and personal reasons. Second, my tweets for #BigData began to increase in
October 2012. At that time, I was knee-deep into the research for my previous
book, Too Big to Ignore: The Business Case for Big Data. Before then, I didn’t tweet
about #BigData very often, much less the title of the book (#TooBigToIgnore).
But not everything changes—at least with me. For instance, my tweets
about #BreakingBad and #Rush have remained fairly constant over time,
with a few notable exceptions. (Did I really go a whole month in early 2013 without
mentioning Canada’s finest export on Twitter?)

* It used to be more open and has recently earned the ire of many developers for allegedly heavy-
handed tactics. For more on the Twitter API, see https://twitter.com/twitterapi.

cintro 5 January 28, 2014 9:50 PM

 6 ▸   B o o k Overview and Background

Figure I.1 Vizify Phil Simon Profile

Image courtesy of Vizify

Figure I.2 Vizify Representation of @philsimon Tweets

Image courtesy of Vizify

cintro 6 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 7

▲
Note
Vizify allows users to customize their public profiles, as well as see their other frequently used
Twitter hashtags. Figure I.2 shows a snapshot of my top-five hashtags as of June 2013.*

Even though this was a one-time experiment, I could see using Vizify on a
regular basis. My tweets will continue to evolve, probably mirroring my pro-
fessional endeavors and newfound personal interests. Case in point: my pub-
lisher has scheduled this book to be released in early 2014. If I run Vizify again
around that time, I would assume that many of my tweets will contain the
hashtag #dataviz. Fortunately, that won’t be difficult to discern.
So, my tweets have changed over time, but (as you can probably tell) this
process just whet my appetite. I was still curious about my Twitter habits, and
other questions remained, like this one: what were my peak tweeting hours?
It took only a few clicks to answer that question. Vizify created a personal
30-second video analyzing my tweets.† A screenshot from that video is pre-
sented in Figure I.3.

Figure I.3 @philsimon Tweets by Hour of Day

Image courtesy of Vizify

* Vizify even lets users create 30-second Twitter videos based on pictures tweeted. To see mine, go
to https://www.vizify.com/phil-simon/twitter-video.
† To see my video, go to https://www.vizify.com/phil-simon/twitter-video.

cintro 7 January 28, 2014 9:50 PM

 8 ▸   B o o k Overview and Background

Since I have always been a morning person, it’s no shocker that my first
tweets start as early as 6 a.m. On most days, I wake up by 5 a.m. and promptly
make myself a cup of coffee. I check my e-mail and tweet new posts or articles
I’ve written for my clients or my own sites. I intentionally break up my normal
day to give my weary eyes a rest by hitting the gym around 10 a.m. By 6 p.m.,
I’ve already put in more than a full day. I’m rarely in front of my computer
after that time, although, like many people, I have recently embraced the two-
screen experience of tweeting when I watch television. (It’s a sign of the times.
For many TV viewers today, “It is a common practice to tweet while watch-
ing. Nielsen has new research that confirms for the first time that tweets can
increase a TV program’s ratings.”1)
Vizify confirmed what I expected: I am not much of a late-night tweeter,
although I occasionally schedule tweets and let HootSuite auto-tweet for me.
(I generally try to space out my tweets, and I don’t follow anyone who tweets
34 times per hour. It’s fair to say that I have developed my own Twitter phi-
losophy. I’d even call myself a bit of a Twitter snob.)
Aside from my most frequently used hashtags, Vizify also identified my most
frequent targets—that is, the people about whom I tweeted most often. I have par-
ticular affection for author and professor Terri Griffith (@terrigriffith) and
blogger Jim Harris (@ocdqblog). In Jim’s case, the feeling is mutual.*
Vizify let me indulge in what was mostly an intellectual exercise. (I can’t
say that my boss forced me to geek out.) I was curious about my tweeting his-
tory and decided to play around with a new toy, hardly an uncommon occur-
rence for me. And there is a slew of other toys. For instance, Ionz lets users
easily “self-visualize” their Twitter data, and Visually lets users do something
similar with their Facebook data. But data visualization is anything but the sole
purview of geeks like me with admittedly too much time on their hands. Social
networks are becoming more interactive, data driven, and visual.
Twitter senior management pays close attention to what its ecosystem and
competition are doing, as it should. Not that Twitter is alone here. For instance,
in June 2013, Facebook added verified accounts, support for hashtags, and
Vine-like video capabilities to its Instagram app. Seem familiar? Facebook
clearly “borrowed” these features from Twitter. Such is life in the Age of the
Platform; frenemies and coopetition are the norm. During that same month,
Twitter added enhanced native analytics of its own.† I have presented my own
in Figure I.4.
I’ll spare you any more analysis of my tweets. You get it. This little yarn
only serves to illustrate one of the key points in this book: it’s never been easier
or more essential to visualize data.

* You can watch Jim’s video here: https://www.vizify.com/jim-harris-1/twitter-video.

† To see yours, just go to http://tinyurl.com/analytics-twitter and log in.

cintro 8 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 9

Figure I.4 @philsimon Twitter Analytics

Source: Twitter

Contemporary Dataviz 101

Incessant social media, memes, and nonstop content permeate our lives. With
seemingly every new hot topic or trend, there’s no shortage of definitions,
many of which come from people and organizations with vested interests in
their definition winning (read: consulting firms, software vendors, and thought
leaders).
In both The Age of the Platform and Too Big to Ignore, I devote a fair amount of
space to defining in plain English my key terms platforms and Big Data, respec-
tively. There’s so much noise and confusion out there on each topic. I feel the
need to do the same here with data visualization.
▲

Note
In this book, contemporary data visualization, or dataviz, signifies the practice of representing
data through visual and often interactive means. An individual dataviz represents information
after it been abstracted in some schematic form. Finally, contemporary data visualization
technologies are capable of incorporating what we now call Big Data.

Primary Objective

There’s a surfeit of data-oriented terms in the business world right now because
data is just plain hot. Let me be absolutely clear here: modern-day dataviz is

cintro 9 January 28, 2014 9:50 PM

 10 ▸   B o o k Overview and Background

not just a synonym or a fancy term for data mining, business intelligence, the
many forms of analytics,* or enterprise reporting.
Delineating among all these terms isn’t terribly important here. Chapter 2
returns to this subject. For now, suffice it to say that these concepts aren’t com-
pletely unrelated to one another. In fact, there’s a great deal of overlap among
them. The most obvious: each is predicated on data in one form or another.
Views on the “proper” goal of dataviz vary considerably. For instance, con-
sider the words of Vitaly Friedman, the editor-in-chief of Smashing Magazine, an
online periodical for professional Web designers and developers:
The main goal of data visualization is its ability to visualize data,
communicating information clearly and effectively. It doesn’t
mean that data visualization needs to look boring to be functional
or extremely sophisticated to look beautiful. To convey ideas
effectively, both aesthetic form and functionality need to go hand
in hand, providing insights into a rather sparse and complex
dataset by communicating its key aspects in a more intuitive way.
Yet designers often tend to discard the balance between design and
function, creating gorgeous data visualizations which fail to serve
its main purpose—communicate information.2

Communicating information is unquestionably important, but not every-

one believes in its primacy (read: that it should be the goal of dataviz). Data-
viz pioneers Fernanda Viégas and Martin M. Wattenberg have suggested that
the ideal dataviz goes beyond promoting understanding and communication.
Those are short-term goals that should buttress the long-term aims of making
better business decisions and even prediction. We’ll see in Part II how Visual
Organizations use dataviz to do much more than understand what’s currently
happening.
Dataviz applications are certainly important, but it’s best to think of data
visualization as more than the output of some software program. So argues
Nathan Yau in his 2013 book Data Points: Visualization That Means Something.
Yau stresses the importance of thinking of dataviz more as a medium than a
specific tool. “Visualization is a way to represent data, an abstraction of the
real world, in the same way that the written word can be used to tell differ-
ent kinds of stories,” he writes. “Newspaper articles aren’t judged on the same
criteria as novels, and data art should be critiqued differently than a business
dashboard.”3
I could quote other dataviz experts ad infinitum, but I won’t belabor the
point: opinions on the topic are far from unanimous. For the purposes of
this book, dataviz shares the same ultimate goal with data mining, business
intelligence (BI), analytics, and enterprise reporting: to make more informed

*T
 hese include standard analytics, Big Data analytics, visual analytics, not to mention industry-
specific analytics like retail, health care, and manufacturing.

cintro 10 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 11

business decisions. Contemporary dataviz is primarily a means of exploring

data and discovering valuable insights. It is not about reporting per se nor is it
about creating pretty graphs or charts for the sake of doing so. In other words,
the most valuable data visualizations today are often based on the premise
that employees don’t knowing exactly what they’re looking for, much less
what they’ll find. By exploring the data, employees are more likely to discover
interesting tidbits or revelations that should drive better decision-making and
outcomes. (You won’t find too many absolutes and guarantees in this book.
I’m a big fan of probabilistic thinking.)

Benefits

To be sure, data doesn’t always need to be visualized, and many data visuali­
zations just plain suck. Look around you. It’s not hard to find truly awful
representations of information. Some work in concept but fail because they
are too busy; they confuse people more than they convey information, to para-
phrase the late George Carlin.* Visualization for the sake of visualization is
unlikely to produce desired results—and this goes double in an era of Big Data.
Bad is still bad, even and especially at a larger scale.
John Sviokla serves as the vice chairman of Diamond Management &
Technology Consultants. As he writes on the Harvard Business Review blog,4
dataviz confers three general benefits:
1. Great visualizations are efficient. They let people look at vast quantities
of data quickly.
2. Visualizations can help analysts or groups achieve more insight into the
nature of a problem and discover new understanding.
3. A great visualization can help create a shared view of a situation and
align folks on needed actions.

At a high level, Sviokla is spot-on. Consider the following example,

as it demonstrates how quickly even a simple dataviz can communicate
information. Figure I.5 shows a comical visual of six prominent companies’
2011 org charts.
Would it be hard to write a few sentences on each organization’s structure?
Of course not. In early 2011, Apple revolved around one iconic man. Even
after Steve Jobs’s death, his presence is deeply felt throughout the company.
For its part, Oracle is still a litigious company. Microsoft is composed of war-
ring factions. Looking at the six images in Figure I.5 represents a quicker and
certainly more humorous way of summarizing each company than even my
witty text probably could.

* You can watch his rant on language here: http://tinyurl.com/carlin-language.

cintro 11 January 28, 2014 9:50 PM

 12 ▸   B o o k Overview and Background

Figure I.5 Organizational Charts (2011)

Source: Manu Cornet

I would add that, for the purposes of understanding large, unpredictable

datasets, interactive data visualizations are generally superior to static infographics,
dashboards, and standard reports. (I should know. I’ve designed thousands of
the latter in my consulting career for my clients.) By definition, presenting even
Small Data in predetermined, static, noninteractive formats limits what users
can do with—and ultimately get from—data. This has always been the case. In
other words, these types of formats generally preclude people from interacting
with the data.* They can’t drill down and around. They can’t explore, nor can
they ask iterative and better questions, and ultimately find answers.

*G
 ood report writers know that it’s not terribly difficult to add some level of interactivity to static
reports. For one example of how to do this, see http://tinyurl.com/phil-crystal.

cintro 12 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 13

More Important Than Ever There are myriad questions

Like all sentient beings, we humans have always that we can ask from data
processed information in different ways, or at least today. As such, it’s impossible
attempted to do so. Many researchers have proven to write enough reports
that there is something unique about how we see or design a functioning
information, as opposed to how we hear it. Many dashboard that takes into
excellent studies and books have informed our cur- account every conceivable
rent understanding of the workings of the human contingency and answers
brain, and I certainly won’t attempt to summarize every possible question.
them all here. The Visual Organization is in no way a
book about the behavioral sciences, neurology, or cognitive psychology. I will,
however, concisely mention a few of the more important works in those fields.
“The human visual system is a pattern seeker of enormous power and
subtlety,” writes Colin Ware in his classic text Information Visualization:
Perception for Design. “The eye and the visual cortex of the brain form a massive
parallel processor that provides the highest bandwidth channel into human
cognitive centers. At higher levels of processing, perception and cognition
are closely interrelated, which is why the words understanding and seeing are
synonymous.”5 Our brains are wired to process information better in a visual
manner.
That humans tend to comprehend visual information quicker than raw
data doesn’t mean that all visualizations are created equal. On the contrary,
we understand certain types of graphical representations better than oth-
ers. Researchers William S. Cleveland and Robert McGill proved as much in
September 1984. Cleveland and McGill published a paper in the Journal of
the American Statistical Association titled “Graphical Perception: Theory, Exper-
imentation, and Application to the Development of Graphical Methods.”
Cleveland and McGill studied the visual clues that people are able to decode
most accurately. The two ranked these clues in the following list:

1. Position along a common scale, e.g., scatter plot

2. Position on identical but nonaligned scales, e.g., multiple scatter plots
3. Length, e.g., bar chart
4. Angle and slope (tie), e.g., pie chart
5. Area, e.g., bubbles
6. Volume, density, and color saturation (tie), e.g., heat map
7. Color hue, e.g., news map

A slightly modified visual of this list is presented in Figure I.6. In

English, it means that not all people comprehend and decode all visual
cues equally. For instance, we tend to understand data positioned along a

cintro 13 January 28, 2014 9:50 PM

 14 ▸   B o o k Overview and Background

Figure I.6 Visual Cues Ranking

Source: Reprinted with permission from The Journal of the American Statistical Association. Copyright 1984
by the American Statistical Association. All rights reserved.

common scale better than data shown on heat maps. Note, however, that the
findings of Cleveland and McGill should not be seen in absolute terms. The
study suggests that absolutes are a myth and that the ability to understand
visual clues is situational. For example, some people will understand a bar
chart better than a bubble chart. The Cleveland and McGill recommenda-
tions are just general guidelines.
In their 2012 book Infographics: The Power of Visual Storytelling, Jason
Lankow, Josh Ritchie, and Ross Crooks demonstrate how even very simple for-
matting can make certain data stand out at the expense of other data. Consider
Figure I.7, a series of random numbers. Go ahead and find each instance of the
number 7.
Now, with simple formatting, repeat the same exercise with Figure I.8.
In professional settings, data has always mattered, although some depart-
ments and industries have been more likely to embrace it than others. In this

Figure I.7 Preattentive Processing Test 1

Source: Lankow, Ritchie, and Crooksa

cintro 14 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 15

Figure I.8 Preattentive Processing Test 2

Source: Lankow, Ritchie, and Crooks

book, I contend that data visualization has never We acquire more information
been more important. Chapter 1 will have a great deal through our visual system
more to say about the rise of the Visual Organization. than we do through all our
For now, suffice it to say that representing informa- other senses combined. We
tion in schematic forms has always been essential to understand things better and
human understanding.
quicker when we see them.

Revenge of the Laggards: The Current State of Dataviz

Fifteen years ago, the presentation of data wasn’t terribly democratic, sophis-
ticated, and interactive, especially compared to today. Tech-savvy analysts and
IT professionals generated static diagrams, graphs, and charts for quarterly or
annual meetings or “special events.” Back then, cutting-edge dataviz wasn’t
part and parcel to many jobs. There just wasn’t that much data, especially
compared to today.
In a way, this was entirely understandable. Yes, the late-1990s saw the
advent of modern enterprise reporting and BI applications adroit at represent-
ing mostly structured data. In most organizations, however, relatively few peo-
ple regularly visualized data, at least not on a regular basis.
My, how times have changed. Now data is everywhere. As I wrote in Too Big
to Ignore, we are living in the era of Big Data, and many things are changing. In
the workplace, let’s focus on two major shifts. First, today it is becoming incum-
bent upon just about every member of a team, group, department, and organiza-
tion to be, at a minimum, comfortable with data. Fewer and fewer knowledge
workers can hide from quantitative analysis. Second, pie charts, bar charts,

cintro 15 January 28, 2014 9:50 PM

 16 ▸   B o o k Overview and Background

and other simple data visualizations of 15 years ago now seem quaint. They
don’t remotely resemble anything that qualifies as contemporary dataviz. More
important, today they often fail to tell the stories that need to be told.
Next, data no longer needs be presented on an occasional or periodic basis.
We are constantly looking at data of all types—a trend that will only intensify
in the coming years. Before our eyes, we are seeing the ability to effectively
present quantitative information in a compelling manner become a professional
sine qua non. Hidden in the petabytes of structured and unstructured data are
key consumer, employee, and organizational insight. If found and unleashed,
those insights would invariably move the needle.
The PwC survey confirmed what I have long assumed. Although notable
exceptions exist, only a minority of organizations and professionals currently do
very much with dataviz. Most enterprises fail to present data in visually compelling
ways. Far too many rely upon old standbys: bar charts, simple graphs, and the
ubiquitous Excel spreadsheet. And their business decisions suffer as a result.
Why the widespread lack of adoption? I’d posit that several factors are
at play here. First, while dataviz is hardly new, the landscape is. Many of the
applications and services detailed in Chapter 2 are recent advents. Second, I
have little doubt that the explosions of dataviz and Big Data left many CXOs
overwhelmed. In this way, dataviz is much like cloud computing. With myriad
options, it’s natural for those in control of the purse strings to ask, “Where do
we even start?”
Next, many organizations suffer from downright ignorance. Many lack
the knowledge that better tools exist, not to mention that enterprises are suc-
cessfully using them. (Hopefully, this book will change that, at least to some
extent.) Then there are organizations whose cultures systematically ignore
data and analysis. I have seen my share of those As such, their employees
generally lack the willingness to try, buy, deploy, and use contemporary data-
viz tools. When corporate fiat, culture, and politics dominate decision-making,
what’s the point of even looking at data?
For these reasons, it should be no surprise that Big Data is still in its infancy.
Brian McKenna tackles this subject in an April 2013 ComputerWeekly article.
About the state of Big Data, he writes that “Analytics firm SAS and SourceMedia
surveyed 339 data-management professionals about their organizations’ use of
▲

Note
The hype around Big Data and, to a lesser extent, dataviz still far exceeds their business
realities. To quote former Notre Dame coach Lou Holtz, “When all is said and done, more
is said than done.” Rather than hem and haw, organizations should recognize the vast
opportunity that the status quo represents. Those that act now can realize significant benefits
that won’t be available to them once their competition wakes up.

cintro 16 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 17

The Current Data on Dataviz

Sadly, most employees—and, by extension, departments and organizations—don’t capitalize

on the massive opportunities presented by Big Data and data visualization. So says consulting
firm PricewaterhouseCoopers (PwC) in its fifth annual Digital IQ Survey (titled “Digital
Conversations and the C-suite”).
In 2013, PwC surveyed 1,108 respondents from 12 countries across a variety of industries.
Respondents were equally split between IT and business leaders. More than 75 percent
worked in organizations with revenues of more than $1 billion.*

Findings
A majority of respondents (62 percent) think that Big Data can provide a competitive
advantage. That’s not exactly surprising, but believing in the power of Big Data is hardly the
same as turning it into actual business insights—and then acting upon them. Nearly the same
number of respondents (58 percent) agreed that moving from data to insight is much easier
said than done.
Only 26 percent of global survey respondents are currently using dataviz. (I suspect that
many of these “forward-thinking” organizations aren’t exactly Google-like in their execution.)
Interestingly, though, adoption—or lack thereof—is not evenly distributed among all
respondents. Specifically, those that reported revenue growth in excess of 5 percent led the
pack—and weren’t letting up. They planned to invest more in data visualization in 2013. The
same can be said of organizations in the top quartile for revenue, profitability, and innovation.
The gap between the dataviz haves and have-nots seems to be growing.

Obstacles
Organizations face four major obstacles with respect to Big Data:

1. They are blind to the importance of visualization.

2. They are investing more in gathering data than analyzing it.
3. They are facing a talent gap.
4. They are struggling with insufficient systems to rapidly process information.

“The amount of information and data that we’re collecting now is truly enormous, [especially]
the volume that is outside the four walls of the organization,” says Anand Rao, principal at
PwC. “Organizations don’t have the right people, they don’t have the right structure in place,
and they’re still struggling with some of the tools and techniques.”6
Rao points out that many organizations do a passable job at looking backward—that is,
“hindsight analysis.” Far fewer, though, predict very well. As we’ll see throughout this book,
dataviz can be useful in this regard.

* Access the entire report at http://tinyurl.com/pwc-dv-2.

cintro 17 January 28, 2014 9:50 PM

 18 ▸   B o o k Overview and Background

data-management technology in December 2012, discovering reality still lags

behind its hype. Only 12 percent of information management professionals are
doing Big Data, according to a recent survey.”7 Remarkably, only 14 percent of
respondents categorized their organizations as “very likely” to begin working
with Big Data in 2014. Nearly one in five responded “not likely at all.”

Book Overview

Big Data is here, leaving many organizations and their employees overwhelmed.
Fortunately, new data-visualization applications are helping enterprises isolate
the signal in the noise.
For instance, through interactive dataviz tools, Netflix discovers trends,
diagnoses technical issues, and unearths obscure yet extraordinarily valuable
customer insights. Employees at Autodesk use a remarkable and interactive
tool that visualizes current and historical employee movement. From this, they
identify potential management issues and see what a corporate reorg really
looks like. Through cutting-edge dataviz, start-up Wedgies instantly serves
up real-time poll results while monitoring poll traction and site issues. The
University of Texas is bringing a visual type of transparency to academia. It
makes unprecedented amounts and sources of institutional data available on
its website. Anyone with the desire and an Internet connection can slice and
dice a mountain of its data in myriad ways. And then there’s eBay. Powerful
data-discovery tools allow its employees to effectively “see” what ebay.com
would look like as a brick-and-mortar store.
These progressive organizations are the exceptions that prove the rule. Most
enterprises are woefully unprepared for Big Data. Far too many erroneously
believe and act like nothing has really changed. As such, they continue to
depend exclusively on reporting stalwarts like Microsoft Excel, static dash-
boards, basic query applications, and even traditional business intelligence
tools. In so doing, they are missing out on the tremendous opportunities that
new data sources and dataviz tools can provide.
Amidst all the hype and confusion surrounding Big Data, though, a new
type of enterprise is emerging: the Visual Organization. An increasing number
of organizations have realized that today’s ever-increasing data streams, vol-
umes, and velocity require new applications. In turn, these new tools promote
a different mind-set—one based upon data discovery and exploration, not on
conventional enterprise reporting. Interactive heat maps, tree maps, and cho-
ropleths promote true data discovery more than static graphs and pie charts.
Today, a growing number of enterprises have turned traditional dataviz on its
head. In their stead, they are embracing new, interactive, and more robust tools
that locate the signals in the noise that is Big Data. As a result, these enterprises
are asking better questions of their data—and making better business decisions.

cintro 18 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 19

The Visual Organization is a largely positive and forward-thinking book. I

focus more on profiling the organizations and employees who get it, not exco-
riating the ones that don’t. (Trust me. There is no shortage of the latter.) Where
warranted, I do attempt to explain the reasons behind certain types of stasis,
dysfunction, and failure. These observations are based upon both my research
for this book and the decade I spent as an enterprise IT consultant. Let me
be clear: my goal here is not to harp on the negative. Rather, I merely want
readers to understand the ways in which Visual Organizations differ from less
progressive enterprises. As Bill Gates once said, “It’s fine to celebrate success,
but it is more important to heed the lessons of failure.”
In the following pages, you’ll meet some amazing companies and people
who recognize the power of Big Data and dataviz. They are pushing the enve-
lope and looking at problems very differently than their data-challenged coun-
terparts. And they are seeing their efforts bear fruit.

Defining the Visual Organization

While useful and informative, many of the texts on data visualization empha-
size theory more than practice. The Visual Organization does not. The forthcom-
ing chapters introduce some fascinating practitioners who regularly visualize
data to understand it, interpret it, and ultimately take action on it. You’ll dis-
cover, as I did in researching this book, that Visual Organizations have moved
well beyond simple charts, graphs, and dashboards that play nice with struc-
tured, transactional data—aka, Small Data.* They are using new tools to make
sense of unstructured data, metadata (data about data), and other emerging
data types and sources. And, as you’ll see, the results are impressive.
▲

Note
Since this is a book about Visual Organizations, a short, formal definition is in order:
A Visual Organization is composed of intelligent people who recognize the power of data.
As such, it routinely uses contemporary, powerful, and interactive dataviz tools to ask better
questions and ultimately make better business decisions. As we’ll see in Chapter 6, the notion
of a Visual Organization is not binary; there are four levels. More advanced enterprises use
interactive data-visualization applications to analyze Big Data. They recognize the inherent
limitations of Small Data and static dataviz.

Central Thesis of This Book

The Visual Organization is based on a simple premise. The Data Deluge has arrived,
and it isn’t going anywhere. More than ever, employees and organizations

* Examples include a list of sales or employees. Think orderly and Excel-friendly data.

cintro 19 January 28, 2014 9:50 PM

 20 ▸   B o o k Overview and Background

have to process and understand unprecedented amounts of information—or at

least try. Complicating matters, new types and sources of data are flying at us
faster than ever. Consider this amazing fact from The Human Face of Big Data, a
fascinating book by Rick Smolan and Jennifer Erwitt. Today the average man
is exposed to more data in a single day than his fifteenth-century counterpart
was in his entire lifetime! According to an oft-cited March 2013 U.N. study,
today more people can access cell phones than toilets.* Out of an estimated
7 billion people on the planet, roughly 6 billion can use mobile phones. Only
4.5 billion can say the same about working commodes.
Alternatively stated, data is streaming at us with increasing variety, velocity,
and volume, with no discernible end in sight. These are the well-documented
three v’s of Big Data. Against this backdrop, intelligent organizations have
realized several things. First, data visualization is becoming essential, and not
just to manage discrete events. Visual Organizations benefit from routinely
visualizing many different types and sources of data. Doing so allows them to
garner a better understanding of what’s happening and why. Equipped with
this knowledge, employees are able to ask better questions and make better
business decisions. As companies like Amazon, Apple, Facebook, Google, Twitter,
Netflix, and others have shown, discoveries from Big Data can represent a huge
competitive advantage. To do this, they have had to buy and build new tools.
Yes, old standbys like Microsoft Excel spreadsheets, charts, dashboards, key
performance indicators, and even mature business intelligence tools still matter.
By themselves, however, they are no longer sufficient to cope with the Data Deluge †.
Bottom line: we live in a world rife with Big
This is not a book about Data. Organizations and their employees need
how to visualize data per se. different applications to find the needles buried in
Rather, it is a book about the haystacks, comprehend immense and dynamic
Visual Organizations. datasets, and ultimately make better business
decisions.

Cui Bono?

In any given month, I typically talk to a wide variety of people: CXOs, con-
sultants, freelancers, mid-level managers, entry-level employees, unemployed
professionals, journalists, fellow authors and speakers, professors, and college
and graduate students. Some live in the United States, others abroad. They
work at organizations that run the gamut: tiny start-ups, small businesses, and
large corporations. And they work for nonprofits, government agencies, and
the private sector. Although the conversations vary, I have noticed a recurring

* To read more, go to http://tinyurl.com/un-toilets.

† Doug Laney of Gartner coined the three v’s in February of 2001. For more on this, see http://

tinyurl.com/gartnervs.

cintro 20 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 21

theme over the past few years: most people are simply overwhelmed by data.
They are struggling to cope with this deluge.
I wrote The Visual Organization for all of these people.
At its core, this book demonstrates how intelligent people and organiza-
tions are making better business decisions via contemporary dataviz new data
visualization applications. Contemporary dataviz is no longer just nice to have
or fodder for quarterly presentations. Organizations are increasingly embracing
new dataviz tools, Big Data, and, most important, a new, data-driven mind-set.
Visual Organizations and their employees are handling the Data Deluge bet-
ter than their “visually challenged” counterparts. Finally, they distinguish
between traditional reporting and data discovery.
In the forthcoming chapters, I’ll demonstrate that dataviz is becoming indis-
pensable, but make no mistake: it is no elixir. It does not solve every conceivable
business problem. No matter how insightful, no matter how much data they
present, data visualizations do not always provide the right answer, much less
guarantee flawless execution. Often a dataviz only serves to clarify an existing
issue, and there’s no guarantee that it will shed light on every possible problem.
Limitations aside, the need for—and power of—dataviz has never been
more pronounced, a fact that the Visual Organizations profiled in this book
and their employees completely understand.

Methodology: Story Matters Here

Of all the companies started around the time of the dot-com boom, Amazon
remains one of its few survivors. Calling it a survivor, however, is the acme of
understatement. The company is nothing short of a titan—the Walmart of the
Internet. And Amazon is causing unexpected ripple effects for a slew of com-
panies and industries.
As I write these words, Oracle and its CEO Larry Ellison are forging partner-
ships with longtime rivals Microsoft and Salesforce.com.* The companies are
putting aside their acrimonious histories with one other. They have struck an
important alliance that attempts to preserve their footholds in the enterprise.
At the core of their newfound and unexpected cooperation: a common fear
of Jeff Bezos’s firm. Amazon is a threat to them all. Ellison, Salesforce.com
head Eric Benioff, and Microsoft big kahuna (at least, as of this writing) Steve
Ballmer clearly understand the old Arabic proverb, “The enemy of my enemy
is my friend.”
Despite Amazon’s longstanding prominence, the purportedly definitive
text on was only recently written. Magazines like Wired have covered different


* The Oracle alliance put Salesforce.com’s cloud-based CRM software atop Oracle apps and infra-
structure. Cats and dogs living together…

cintro 21 January 28, 2014 9:50 PM

 22 ▸   B o o k Overview and Background

aspects of the company very well and in some depth. Nearly 20 years after its
founding, Amazon lacks the equivalent of an authorized tell-all, a compre­
hensive window into its vast business. Up until recently, the books written
about the company have been at best incomplete and at worst disappointing.
That finally changed in October 2013. Bloomberg Businessweek reporter Brad
Stone published his much-anticipated book The Everything Store: Jeff Bezos and
the Age of Amazon. The book is the closest thing available to a comprehensive
company biography. Stone interviewed hundreds of former executives and
operations, and I have eagerly followed the status of his book since it was
announced.
As an author and occasional journalist, I am familiar with these types
of press-related obstacles. (Maybe privacy isn’t completely dead after all.)
In researching previous books, I have contacted folks at high-profile com-
panies, some of whom I would even call friends. My requests to speak on
the record to employees in the know were politely denied, whether it was
about privacy at Google or about Big Data at Facebook. In each case, these
folks kindly told me that, as much as they may want to help me, their
employers took controlling the message very seriously. I was disappointed
but not offended. I understood. Access to senior management about propri-
etary or sensitive subjects isn’t easy to come by, especially if the result is a
book or an article.
None of this should be surprising. Steve Jobs only agreed to an autho-
rized biography with Walter Isaacson when the former faced his imminent
mortality. For years Jobs denied requests by authors and publishers to do
the same thing. Like Jobs, Bezos is by all accounts a very private person,
and Amazon follows the lead of its iconic CEO. Letting journalists and
authors into their walled gardens ultimately serves no real business pur-
pose. The risks far outweigh the rewards. Companies on that level aren’t
exactly hurting for PR, and flying under the radar suits them just fine.
Sanctioned books like In the Plex: How Google Thinks, Works, and Shapes Our
Lives by Steven Levy are the exceptions that prove the rule. (Levy’s access
to Google was unprecedented.)
In 2009, the AMC Network launched a new slogan: Story Matters Here. I
couldn’t agree more. For a book like The Visual Organization to work, I would
have to do a good bit of research. That meant identifying organizations visu-
alizing their data in interesting ways, making better business decisions as a
result. In the Internet age, I knew that that wouldn’t be terribly hard to do.
Aside from my personal connections, I could use Google, Facebook, Twitter,
LinkedIn, and other indispensable sites for research purposes.
But that wasn’t all. To do this book right, I needed to do two other things.
First, I would have to find dataviz practitioners doing cutting-edge things—and

cintro 22 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 23

then talk to them. The dozens of conversations I had with dataviz professionals
inform the pages that follow, whether or not I ultimately profiled their compa-
nies and clients. I learned a great deal, as I hope you will.
I have always aspired to write more books that are more “show me don’t
tell me.” To that end, I knew that the book’s case studies would be key—and I
set that bar relatively high. I was clear with interviewees from the get-go. For
their organizations to be featured in the text, they would have to get specific. I
would need them to provide actual examples of the dataviz tools they used to
do their jobs. Platitudes just wouldn’t cut it.
Now, this wasn’t my first rodeo. I knew that my self-imposed second
requirement would pose more challenges than my first. For instance, United
Parcel Service uses technology and data in truly amazing ways. UPS routes its
trucks to millions of homes and businesses each day in an efficient manner.
This process requires incredibly sophisticated algorithms. I am quite certain
that the company’s use of data visualization is book-worthy. In June of 2013,
I reached out to a friend of mine, a UPS employee for more than two decades.
My friend told me exactly what I expected: UPS keeps a low profile and does
not like to be featured in magazine articles and books. Including fresh UPS
material and examples of its proprietary tools in this book would require
approval at the highest level of the company. (Can someone say lawyers?)
Unfortunately, my efforts to include UPS went nowhere, as did similar
attempts to pioneer new dataviz research on universally recognizable orga-
nizations like the National Basketball Association, Facebook, Twitter, ESPN,
Pandora, and a few others. Just because these companies are not profiled in
The Visual Organization doesn’t mean that they’re not doing fascinating things
with data and dataviz.
With the exception of Netflix, the case studies in Part II meet both of my
two criteria. (I’ll explain the reasons for the slightly different methodology
for Netflix in Chapter 3.) Profiling only relatively forthcoming organizations
with remarkable dataviz stories has resulted in a better book. Such examples
will, I hope, teach the reader important lessons about the subject, including
what to do, what not to do, how to do it, and more. To me, a “story-centric”
approach just made sense. It is superior to one that emphasized company
notoriety at the expense of specifics and transparency. In the end, I believe
that how, why, and what are more important than who. The Visual Organiza-
tion benefits from profiling organizations with compelling and specific exam-
ples of contemporary data visualization, even if a few of those organizations
aren’t necessarily household names. And, as I’ll argue in the following pages,
these lesser-known enterprises may well become more recognized and suc-
cessful precisely because they understand the tremendous value that data and
dataviz offer.

cintro 23 January 28, 2014 9:50 PM

 24 ▸   B o o k Overview and Background

The Quest for Knowledge and Case Studies

One day in August 2013, a graduate student by the name of James Eichinger
tweeted at me. @Ikejames101 is studying predictive analytics at Northwestern
University. In a subsequent e-mail, Eichinger informed me that wanted to do
his final project on data visualization, but he was encountering a major prob-
lem. In his words, “Most of the case studies [he] found are either weak or
tangential to the subject.”8 As for the blog posts on sites like Harvard Business
Review, “Not a single one [shows] impacts on business decisions, management
culture, or information efficiency.” I didn’t entirely concur with Eichinger’s
assessment, but our exchange piqued my curiosity about the prevalence of
proper data-visualization case studies on the Web. I slept on it.
The next morning, I performed three specific Google searches. I queried exist-
ing case studies related to three different types of major enterprise technologies.
The results are presented in Table I.1, and is displayed graphically in Figure I.9.
Even though Table I.1 and Figure I.9 confirmed my suspicions, it should
not be taken as gospel or proof that the dataviz case study landscape is entirely
barren. For three reasons, I wouldn’t go that far.
First, think of Google’s search data here as a proxy of sorts.* Without
carping over the proper definition of the term case study, I have no doubt that
there are more than 23 dataviz stories on the Web. (How many of them are
actually good, useful, and vendor neutral is another matter altogether.)
Second, quantity should never be mistaken for quality. Many of the ERP
and CRM case studies on the Web aren’t terribly instructive. Third, by default,
Google provides increasingly personalized results based upon factors like user
geography, known demographic information, individual browsing history, and
others.† Sometimes identical Google searches from ostensibly similar users
yield wildly different results.

Table I.1 Google Search Results on Three Different Types of Case Studies

Google Search Term

(with Quotes) Approximate Results Notes
“Data Visualization case studies” 23 Interestingly, two of the 23
results came from www.philsimon
.com
“ERP case studies” 6,670 Enterprise resource planning
“CRM case studies” 16,500 Customer relationship management
Source: Google, as of August 31, 2013


* As any experienced Googler knows, small changes in search terms can yield vastly different results.
†
Users can easily turn this “feature” off if they like.

cintro 24 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 25

20,000

15,000

10,000

5,000

0
“Data
Visualization
“ERP case
case studies”
studies” “CRM case
studies”

Figure I.9 Graph of Google Search Results on Three Different Types of Case Studies
Data Source: Google, as of August 31, 2013

Despite these qualifications, the differences in my little case study experi-

ment are irrefutable. The number of CRM and ERP profiles are larger than their
dataviz counterparts by orders of magnitude (717 and 290 times, respectively).
Put differently, Table I.1. and Figure I.9 only illustrate what I, Eichinger, and
countless others have found: profiles of organizations using contemporary
dataviz and new forms of data not written by marketing departments are lack-
ing. Data visualization is becoming critical, but how have organizations are
really doing it? And what are the lessons? What are the pitfalls?
And that’s where The Visual Organization steps in. It is my sincere hope
that the observations, framework, case studies, interpretation, and original
research in this book will help organizations use dataviz to move their needles.
Academic studies and journal articles are certainly beneficial, but as of this
writing there is a paucity of vendor-neutral case studies on the subject. Along
these lines, perhaps this book will also make a meaningful contribution to the
field of data visualization.

Differentiation: A Note on Other Dataviz Texts

The Visual Organization is hardly the first book about dataviz. On the contrary,
many other researchers, authors, and practitioners have contributed a great
deal to the field. It’s no understatement to say that a vast body of work has
been done on the topic.
In their books, Stephen Few, Edward Tufte, Alberto Cairo, Colin Ware, and
Nathan Yau explain how to effectively visualize data very well. They cover
the mechanics of creating graphs, charts, and, more recently, infographics,

cintro 25 January 28, 2014 9:50 PM

 26 ▸   B o o k Overview and Background

heat maps, tree maps, and choropleths. Many of these authors’ books illus-
trate best design practices and serve as how-to guides, and I recommend
checking them out. For their parts, dataviz researchers like Marek Walczak,
Martin M. Wattenberg, and Fernanda Viégas have gone way beyond extend-
ing our current understanding of dataviz. They have created exciting new
ways to visualize data. The Visual Organization does not attempt to replicate
their work here.
Nor is this a text primarily about how the human brain processes data. I
don’t cover the science behind the mind’s ability to understand information
represented in a visual form. I’m the furthest thing from a neurologist. Again,
a panoply of excellent books has already been written on the subject. The
bottom line, as data journalist John Burn-Murdoch writes in The Guardian, is
that “Humans are visual creatures. Peer-reviewed studies have shown that we
can consume information more quickly when it is expressed in diagrams than
when it is presented as text.”9
The Visual Organization demonstrates how and why a growing number of
organizations are visualizing their data to diagnose issues, discover new cus-
tomer insights, and make better decisions.

Plan of Attack

The Visual Organization consists of four parts. Part I, “Book Overview and Back-
ground,” examines the reasons behind the ascent of the Visual Organization.
It also covers the five general categories of contemporary dataviz applications
and services.
Part II, “Introducing the Visual Organization,” introduces a number of
diverse Visual Organizations. You’ll discover how Netflix, Wedgies, Autodesk,
and other enterprises have embraced Big Data and dataviz, and not just as dis-
crete one-time “projects.” We’ll see how Visual Organizations have garnered
profound customer insights and solved thorny business problems through new
dataviz techniques and applications.
Part III, “Getting Started: Becoming a Visual Organization,” takes a step
back. It begins by providing a framework for readers to understand the four
different levels of Visual Organizations. It then asks a key question before
extrapolating a series of lessons, best practices, myths, and mistakes from the
case studies in Part II. No, it’s not a checklist to follow for becoming a Visual
Organization, but it does present sage advice for readers interested in both
reaping the benefits of dataviz and avoiding their common pitfalls.
Part IV, “Conclusion and the Future of Dataviz,” concludes the book. It
offers a number of careful predictions about current trends, Visual Organiza-
tions, Big Data, and the future of data visualization.

cintro 26 January 28, 2014 9:50 PM

 I n t r o d u c t i o n    ◂ 27

Next

Chapter 1 examines the ascent of the Visual Organization. It explains what’s

happening and why. We’ll soon see that important business, cultural, tech-
nological, and human shifts are collectively causing enterprises of all kinds to
change the way they think about data and traditional reporting.

Notes

1. Snider, Mike, “Twitter Can Boost TV Ratings,” USA Today, August 6, 2013,
http://www.usatoday.com/story/tech/personal/2013/08/06/nielsen-
twitter-affects-tv-ratings/2613267, Retrieved August 27, 2013.
2. Friedman, Vitaly, “Data Visualization and Infographics,” Smashing Magazine,
January 14, 2008, http://www.smashingmagazine.com/2008/01/14/monday-
inspiration-data-visualization-and-infographics, Retrieved June 12, 2013.
3. Yau, Nathan, Data Points: Visualization That Means Something, Hoboken, NJ:
Wiley, 2013.
4. Sviokla, John, “Swimming in Data? Three Benefits of Visualization,” Harvard
Business Review Blog Network, December 4, 2009. http://blogs.hbr.org/
sviokla/2009/12/swimming_in_data_three_benefit.html, June 11, 2013.
5. Ware, Colin, Information Visualization: Perception for Design, Morgan
Kaufmann, 2000.
6. Olavsrud, Thor, “4 Barriers Stand Between You and Big Data Insight,”
CIO.com, April 9, 2013, http://www.cio.com/article/731503/4_Barriers_
Stand_Between_You_and_Big_Data_Insight, Retrieved August 27, 2013.
7. McKenna, Brian, “SAS: Data Quality, Data Governance Concerns Impede
Big Data Programmes,” ComputerWeekly.com April 3, 2013, http://www
.computerweekly.com/news/2240180600/SAS-data-quality-data-gover-
nance-concerns-impede-big-data-programmes, Retrieved August 30, 2013.
8. E-mail from Eichinger, August 31, 2013.
9. Burn-Murdoch, John, “Why You Should Never Trust a Data Visualisation,”
theguardian.com, July 24, 2013, http://www.guardian.co.uk/news/
datablog/2013/jul/24/why-you-should-never-trust-a-data-visualisation,
Retrieved July 24, 2013.

cintro 27 January 28, 2014 9:50 PM