I visited the Sydney Museum of Contemporary Art yesterday. Among the many works—some marvelous, some not so much—one in particular caught my attention. It was titled something like “My life is nuts.” Here it is:

This piece of art consists of a peanut for everyday that the artist has lived. It is a “work in progress” in that each day the museum staff adds another nut to the pile.

Because I’ve been thinking a lot about so-called Big Data these days, I couldn’t help but see a connection between Big Data and this pile of peanuts. Just as this pile of nuts is an incremental representation of the artist’s days on earth, Big Data is an incremental increase in data volume, velocity, and variety. Two differences exist, however. The Big Data pile of nuts increases at an exponential rate, unlike the linear increase of the artists lifespan, and the artist hasn’t chosen to give herself a new name for arbitrary points along the growth of this pile. Data has been increasing at an exponential rate since the advent of the computer; it didn’t suddenly become big and it hasn’t passed some threshold that makes is qualitatively different than it was in the past.

Last Thursday I spoke at the ActuateOne Live event in San Francisco. Immediately before my keynote, Shaku Atre, a longtime thought leader in the field of business intelligence with a resume that dates back at least as far as mine spoke about the “Top Ten Rules of Big Data Systems.” When I listened to her presentation, I expected to hear opinions about Big Data that differed from mine, but I was surprised to find none. I’m sure that our opinions differ in some respects, but this wasn’t obvious from the content of her presentation. I pointed out to her afterwards that the contents of her presentation could have been titled “Top Ten Rules of Business Intelligence Systems” and given ten years ago with little alteration. She didn’t disagree. She said, “You are right—I seem to see Big Data as a fluid continuation of the past rather than a significant departure from the past, as you do.” Anticipating the allure of all things big, bigger, and biggest, Atre has already acquired the rights to the URL “humungousdata.com” and she led a panel discussion at the event titled “Humungous Data? No problem with business analytics.” In cautionary contrast, one of the slides in my presentation included the words “Big Data, little information.”

Jesus once taught a lesson using the parable of the wine skins.

No one pours new wine into old wineskins. If he does, the wine will burst the skins, and both the wine and the wineskins will be ruined. No, he pours new wine into new wineskins.

The fermentation of new wine requires a vessel that can expand. New wineskins were flexible, but old wineskins no longer had the capacity for expansion. Jesus was saying that his message was new—a departure from the past—and that it could not be contained within the paradigms of the past. It was qualitatively different, and as such, following him would require a new perspective.

I’m trying to teach the lesson that’s on the flip side of this parable. Only marketers (people who are trying to sell you something) pour old wine into new wineskins. Doing so wastes the capacity of the new wineskins, because expansion is no longer needed, and it misleads buyers into thinking that they’re getting new wine that is actually old. Big Data isn’t new. The purpose of the term Big Data as it’s being used by technology vendors and most technology thought leaders is to create an illusion of newness that potential buyers can’t live without. It is a marketing campaign designed to lighten the wallets of organizations. An organization that cannot derive value from the data that it already has will not suddenly derive value from it by installing the latest technology. The business intelligence industry (and now the Big Data industry) has always been good at making promises that it rarely fulfills.

Big Data is just a big, exponentially growing pile of nuts. Value can be derived from data, regardless of size, but only to the degree that you’ve developed data sensemaking skills. Only then will Big Data lead to Big Information.

Take care,

I found a kindred spirit when I recently read Nate Silver’s new book The Signal and the Noise (Penguin Press, 2012). I want to give you a sense of the book and it’s powerful message by sharing a few excerpts from the introduction.

This is a book about information, technology, and scientific progress. This is a book about competition, free markets, and the evolution of ideas. This is a book about the things that make us smarter than any computer, and a book about human error. This is a book about how we learn, one step at a time, to come to knowledge of the objective world, and why we sometimes take a step back.

This is a book about prediction, which sits at the intersection of all these things. It is a study of why some predictions succeed and why some fail. My hope is that we might gain a little more insight into planning out futures and become a little less likely to repeat our mistakes.

He talks about the greatest revolution in information technology since the invention of writing—not the so-called information age of today, but the invention of the printing press in 1440.

The amount of information was increasing much more rapidly than our understanding of what to do with it, or our ability to differentiate the useful information from the mistruths. Paradoxically, the result of having so much more shared knowledge was increasing isolation along national and religious lines. The instinctual shortcut that we take when we have “too much information” is to engage with it selectively, picking out the parts we like and ignoring the remainder, making allies with those who have made the same choices and enemies of the rest.

…

But somehow in the midst of this, the printing press was starting to produce scientific and literary progress. Galileo was sharing his (censored) ideas, and Shakespeare was producing his plays.

…

“[But] men may construe things after their fashion / Clean from the purpose of the things themselves,” Shakespeare warns us through the voice of Cicero—good advice for anyone seeking to pluck through their newfound wealth of information. It was hard to tell the signal from the noise. The story the data tells us is often the one we’d like to hear, and we usually make sure that it has a happy ending.

Silver goes on to describe the impact of the printing press over the next few centuries, and concludes with this note of caution:

The explosion of information produced by the printing press had done us a world of good, it turned out. It had just taken 330 years—and millions dead in battlefields around Europe—for those advantages to take hold.

Here’s the clincher:

We face danger whenever information growth outpaces our understanding of how to process it. The last forty years of human history imply that it can still take a long time to translate information into useful knowledge, and that if we are not careful, we may take a step back in the meantime.

I’ve been working to squeeze value from information technology for 30 of these last 40 years. I’ve watched with great dismay as we’ve taken many steps back and want to scream as I see this happening to an unprecedented degree under the banner of Big Data. Progress can be made—the opportunity is ripe for the plucking—but the answers do not lie in the solutions that BI vendors are selling; they lie within us.

Silver goes on:

The exponential growth in information is sometimes seen as a cure-all, as computers were in the 1970s. Chris Anderson, the editor of Wired magazine, wrote in 2008 that the sheer volume of data would obviate the need for theory, and even the scientific method.

This is an emphatically pro-science and pro-technology book, and I think of it as a very optimistic one. But it argues that these views are badly mistaken. The numbers have no way of speaking for themselves. We speak for them.

…

Data-driven predictions can succeed—and they can fail. It is when we deny our role in the process that the odds of failure rise. Before we demand more of our data, we need to demand more of ourselves.

And here’s one final excerpt from the book’s introduction:

Big Data will produce progress—eventually. How quickly it does, and whether we regress in the meantime, will depend on us.

…

Meanwhile, if the quantity of information is increasing by 2.5 quintillion bytes per day, the amount of useful information almost certainly isn’t. Most of it is just noise, and the noise is increasing faster than the signal. There are so many hypotheses to test, so many data sets to mine—but a relatively constant amount of objective truth.

…

The signal is the truth. The noise is what distracts us from the truth.

After reading these words in the book’s introduction, you can imagine how excited I was to dive headlong into the book. It was time well spent.

Some chapters interested me more than others (I care little about baseball), but each illustrates revealing failures and successes of prediction. The one disappointment for me was the chapter on climate change. This falls outside of Silver’s expertise and I’m assured by colleagues who know climate science well that Silver didn’t get a balanced perspective when he relied on others to bring him up to speed, including one fellow in particular who is paid by powerful interests to cast misleading doubt on the merits of climate models.

If you’re looking for a book that will teach you how to develop reliable predictive models, you’ll need to look elsewhere. This is not a how-to book. Silver points to Bayesian thinking as the approach that will inform and continually improve good predictive models, but he provides only a rough introduction to this approach. This is intentional. The skills that are required to build good predictive models cannot be learned from a few chapters. The Signal and the Noise will raise awareness and point the way to better predictions. It is up to us to develop the skills to make this happen.

Take care,

Big Data is a marketing campaign. Uttered from the lips of technology companies, including analyst organizations such as Gartner, the term Big Data remains ill-defined. This is intentional. It allows them to claim just about anything they want because their claims can’t be fact-checked if you don’t actually know what Big Data is.

Generally speaking, technology companies use the term Big Data to refer to greater volumes and new sources of data. This, however, is not a new thing. Since the advent of computers, each year we’ve accumulated more data and new data sources. Data didn’t suddenly become big. Big Data is just more of the same, but it is celebrated by technology vendors, analyst groups, and thought leaders as a qualitative break from the past—the newest techno-panacea that everyone must invest in or be left behind. Claims of Big Data’s effects on the world are growing ever larger and more absurd.

In his keynote presentation at Gartner’s IT Expo last week in Orlando, SVP and Global Head of Research Peter Sondergard proclaimed that by the year 2015 a total of 4.4 million jobs will be created worldwide to support Big Data. Not only that, but every new Big Data role in the U.S. (1.9 million by 2015) will create jobs for three more people outside of IT. What does this actually mean? What constitutes a Big Data role? Because the term is so vaguely defined, Gartner can claim that any new job in an IT department or for people who work elsewhere with data in any way is in fact a new Big Data position. This is pure fantasy.

And how did Gartner come up with the 4.4 million job figure? My guess is that, after a long night of drinking, they gathered around the Ouija board and let the spirits (that is, their own drunken imaginations) lead them to the answer.

Notice the irony. Here is an organization of industry analysts talking about analytical technology that is engaging in analytical nonsense. No qualified data analyst would make such absurd and groundless predictions. Either the so-called analysts at Gartner have never been trained in data analysis or they are fabricating predictions that serve their own financial interests. Most likely, it’s both. CIOs who buy into these prognostications are either naïve or are, like Gartner, motivated by self-interest. After all, chasing the latest technology is what keeps CIOs employed.

Organizations all over the world rely on groups such as Gartner to guide their IT investments. Are they getting objective and reliable advice? Far from it. Gartner has no incentive to discourage organizations from investing in IT. They make their money by keeping us convinced that we can’t live without the latest technologies, regardless of whether they’re actually needed or actually work. The truth is, analyst organizations such as Gartner are in bed with the very technology vendors whose work they supposedly monitor and critique. They’re having a wild orgy in that bed, rolling in cash, but it is only the end users who are getting screwed. Essentially, Gartner and the like operate as extensions of technology company marketing departments. Gartner is creating demand for its clients’ products and services (yes, the very technology companies that these analyst organizations monitor—supposedly in an objective manner—are their clients, who pay dearly for their support). These products and services aren’t usually needed, they are often ineffective, and in the case of Big Data, they’re ephemeral. Have you noticed that every business intelligence vendor has suddenly become the leading Big Data company without changing anything that they do? Just slap a new name on business as usual and you can get the world to line up at your door.

Look past the marketing hype for analytical (data sensemaking) products that actually work. It doesn’t matter whether they’re called Big Data, analytics, or just plain data analysis tools. What matters is that they help you find the signals that exist in the midst of all that noise in your data and make it possible for you to understand those signals and use that understanding to work smarter than before. Demand that vendors show you how their tools can be used to glean real value from your own data. Ignore their claims and demand evidence. Make them show you how you can make better decisions using their products and services. Unless they can provide that, you don’t need what they’re selling.

Take care,

People who don’t know how to manage performance or merely don’t care tend to prefer dashboards (monitoring displays) that say little and do so poorly. This way they’re never told anything they don’t understand; they’re never forced to get off their butts; they’re never faced with a decision they can’t or don’t wish to handle. For such people, the illusion of control is not only sufficient, it’s preferred over real control.

Idiot lights and flashy dashboard gauges are great if you don’t know what you’re doing and don’t care to learn. Idiot lights are those alerts that light up on your car dashboard to tell you that something’s wrong. They assume, usually correctly, that you don’t understand how the car works and couldn’t handle more information beyond “There’s a problem with the engine; take me to the repair shop pronto.”

Most dashboard gauges are designed to look just like speedometers, fuel gauges, temperature gauges, etc., down to the annoying glare of light on glass.

Do you know how much research went into determining that idiot lights and gauges that look just like those in our cars are the best way to display information on a dashboard for monitoring your organization’s performance? The answer is zilch; none whatsoever. Back in the beginning when we started calling computer-based monitoring displays dashboards, someone had the bright idea of making display widgets that looked like those in cars. This is an example of taking a metaphor too literally. In the early days of cars, some of them included holders for horse whips, even though they were no longer needed. This seems absurd to us now, but it’s no more absurd than assuming that information displays for monitoring the performance of your business should look like gauges on a car. The part of the “dashboard” metaphor that works is the similarity in function between car dashboard gauges, which we use to monitor information about the car and our driving, and monitoring dashboards, which we use to monitor information about the organization’s performance. It is meaningless and downright absurd to stretch the metaphor any further.

There’s a term for aspects of design that emulate old-fashioned, physical objects that were used in a different context: skeuomorphs. In a recent article in Wired titled “Clive Thompson: Retro Design is Crippling Innovation,” Thompson, an award-winning technology journalist, bemoaned the prevalence of skeuomorphs in modern software. Here’s an excerpt:

Despite being lauded for design, Apple is the reigning champion in this field, producing a conga line of skeuomorphs that are by turns baffling and annoying. Its iPhone app, Find My Friends, includes astonishingly ugly, faux stitched leather that wastes screen space. On the new iCal for the Macintosh, things are odder yet: when you page forward, the sheet for the previous month rips off and floats away, an animation so artless you’d swear it was designed personally by Bill Gates.

As a Mac user, I too find paging through iCal annoying as hell. Thompson ends the article by completing the thought above:

And if you really need to flip paper pages on your calendar? Buy a handmade one—and hey, get some nice-quality pencils. Let paper work like paper and screens like screens.

Skeuomorphs aren’t all bad. There are times when they can be used, as N. Katherine Hayles explains, as “threshold devices, smoothing the transition between one conceptual constellation and another.” You might argue that by making dashboard gauges look like those on cars, we’re making it easy for people to use performance monitoring displays. It is certainly true that experience driving a car has already taught you how to read them. That would be a good argument if it took days or even hours to learn how to read more informative forms of display, but this is hardly the case. For example, a bullet graph, which was specifically designed for performance monitoring dashboards takes approximately one minute to learn how to read. One minute of learning is a small price to pay for a lifetime of better information more efficiently acquired.

If you’re good at your job and care about it, you want to be fully informed. You’re not an idiot, so don’t put up with dashboards that treat you like one. Demand the information that you need displayed to the level of your ability.

Yesterday morning I had a Big Data experience when I visited the Louvre Museum in Paris. Within 15 minutes of arriving I was ready to run screaming from the glass pyramid. Why? Because I was overwhelmed. Room after room of artistic works totaling in the hundreds of thousands, each magnificent, was too much. I found it nearly impossible to appreciate a single piece when I was surrounded by so many others. Pick any one of those glorious works and place it before me in a quiet room with good lighting and I would study it for hours. Give me something to read that describes the work—the artist, the medium from which it was created, the historical context—and I would appreciate it for an entire day. Place it among thousands of its brethren and I might fail to see it altogether.

We are surrounded by data. In our present day of so-called Big Data, there is more and more of it every day. Anyone who has ever actually worked with data in an effort to make better decisions knows that most of the data that surrounds us is noise. It’s useless. We seek the signals that reside here and there in the midst of the noise. While I stood there in the Louvre this morning, every piece of art was a masterpiece in its own right—every piece a signal—but to me they were all noise because there was too much for my senses to take in or my brain to fathom. Yes, even signals become noise when we’re overwhelmed. I tried desperately to fix my attention on a single piece, but over and over again I failed. I couldn’t shut out the other voices constantly invading my senses yelling “Look at me!”  Yes, I saw the Mona Lisa with her enigmatic smile from behind the barrier while being jostled by the photo-taking crowd, but I couldn’t connect with her or the genius of da Vinci, whose work I so admire.

Others in the Louvre yesterday added the Mona Lisa and a host of other works to their lists of important encounters and no doubt felt enlarged by their experience. But were they enriched in a meaningful way? Did the artists’ voices reach their ears through the din? Were they awakened or did they merely roll over in their sleep?

Data becomes information only when it informs. For that to happen, we must find a signal in the midst of the noise and study it closely enough to understand it. This takes time. This takes attention. This takes skill. Only when this occurs has something useful entered our minds.

Take care,