In mid-September I had the opportunity to attend the Tableau Customer Conference 2013, held this year in Washington, D.C. It was an excellent event with three motivating days of learning about and connecting with the present, future, and state-of-the-art of data analytics and visualization. Though many people may not yet have heard of Tableau, the analysis and visualization work produced by well-known users of the software is beginning to saturate the media we regularly consume. You can see Tableau in action at Cars.com, The Wall Street Journal, Geekwire, CBS Sports, Gizmodo, and Mashable, as just a few examples.
The work those and others are doing is turning raw data into analysis, and then using that analysis to develop a context-rich “story” able to convey real meaning and insight by simultaneously using both the underlying data and its visual presentation to great effect. Data visualization is particularly compelling because it can (when used effectively) present complex ideas or relationships more simply and in a fashion in which the human mind prefers—visually, so that we can easily see patterns and outliers that are unique.
We will wrap up 2013 with “big data” as the dominant business and tech buzzword of the year. Marketing slogans have been created or revamped to incorporate the seeming imminence of Big Data, and big players to small-timers have jumped on the Big Data bandwagon. Unfortunately for many, these marketing slogans miss the much more relevant attribute of data analysis today. Big data is not new. The collection and analysis of very large data sets is even mandated by the United States Constitution in Article I, Section 2 for the decennial census. It is our increasingly advanced analysis of data—big or otherwise—that has dramatically evolved. Very simply, the tools that we use have advanced spectacularly: The computer, graphical user interface, and software. The result of this revolution is much less in regard to the “bigness” of a data set; rather we are at the forefront of a true transformation of how we consume data visually and interact with data dynamically.
Business data is quickly evolving beyond static PowerPoint decks toward interactive and visually rich dashboards. This was a common thread across all of the interactions, side conversations, demonstrations, and key notes at this data conference for data junkies. But, before we get ahead of ourselves, let’s address a fundamental issue at the core of data analysis: Shouldn’t data speak for itself?
Time to deploy another useful tool in story-telling: the analogy. Imagine if you will, a data set that is made up of the discrete letters and punctuation (including spaces) of the English alphabet. Allow me to demonstrate by throwing a bunch of this English alphabet data on the page, delimited by spaces, like this:– “ ” , . a a a a a a c c d e e e e e f f G h h I I I k l m M n n n o o o o p r r r r s s s s s t t t t t t t T u u u w y y y _ _ _ _ _ _ _ _ _ _ _ _ _
Does this “data” speak for itself? As a matter of fact, it has no meaning beyond the very simple point that you know that there is something on the page. That “something” could be random, or there could be a pattern, and there could even be some meaning. This is perhaps not much different than gazing into a bowl of alphabet soup. Without an effective way to visualize this data set it will be very unlikely that we could successfully analyze and “decode” it beyond identifying basic patterns that in themselves have no value. We cannot arrive at any meaning that has value by simply looking at the data.
I could attempt to be a bit cleverer and create a table to organize this data. Each column could be one bit of data (a, b, c, etc.). I could then show a count of each bit of data in a row, revealing frequency. I could splice the data another way to see if position in relation to other letters shows any discernible pattern that could help me to unlock the mystery of the story behind this data set. But, simple slicing and dicing will not advance my analysis very far. This appears to be just data—even when it is massaged a bit, it does not speak for itself. This experience is not unlike what many organizations must work to overcome in their own data. Marketers, operations managers, and the like cannot simply stare into the alphabet soup of their customer or logistics data. They need to decipher and decode it to develop the story of why their business is performing as it is.
If I had the use of a contextually accurate and effective computing tool that could help me to parse my alphabet data set—and to present it visually—my analysis might reveal something like this:“Get your facts first, then you can distort them as you please.” –Mark Twain
In this rather literal analogy the raw data (letters and punctuation in the English language) did not, and could not, speak for itself. But, with the use of a specialized tool, I could organize the data in a way that structured it visually and helped to reveal meaning. Businesses collect data not because it is, unto itself, a fun or inexpensive thing to do. Businesses are also not required to collect the vast majority of data they collect. For example, accounting and regulatory data accounts for a small percentage of the total amount of data collected by most companies. Businesses collect data in order to extract patterns, analyze, and learn the story of what is driving (or hindering) their success.
Most of the data a company collects is about their customers’ interaction with the company and how it operates. By parsing data in a way that enables visual analysis we can help to unlock a meaningful story behind it. A company can better understand the story associated with customer behavior, for example, than the plain facts that the data alone may present. In that customer behavior story a business can develop context and knowledge which is the crucial step necessary for useful analysis that can answer the question of “why?”.
The alphabet-as-data analogy serves to illustrate the value that can be derived from using advanced tools for data analysis and visualization. There is no inherent value in “big data”. There is no threshold of scale beyond which data will speak for itself. In actual practice, the larger a data set grows to be the more overwhelmed its human interpreters often become, and the greater the need for a way to process it into something consumable. We are at the tipping point of visualized and interactive data sharing—dynamic views are replacing static slides and doing a better job of conveying useful information.
Visualization provides both the underlying data—rather like a built-in proof of concept—as well as the extrapolated meaning of that data. It is the methodology needed for telling a business story. The value of this approach will continue to expand as our ability to process and contextualize data becomes ever more sophisticated. The clever business will not care to think “big” when it comes to data, but will rather think “smart”. Smart data is effectively visualized, contextual, relevant, and meaningful. It tells a story to answer a key question of the business and helps to address the question of “why?” and inform the action of “how”.