Who Can Deny The Power Of Data Visualisation?

Visualising Data

Who hasn’t drawn a simple chart on paper to make a point? Instantly, we focus on it, and as simplistic as it may be it clarifies our minds. It’s no surprise, then, that visualisation continues to become ever more important to business people, no matter what their level or function. It is, in fact, the newest and most critical bridge we have from data to insight to action.

The reasons seem obvious: It’s the most efficient mechanism we’ve ever known that delivers or provokes insight. While spreadsheets require continual scanning and rescanning as we pick up a few numbers at a time, visualised data goes direct. It seems to fly straight into the brain. It’s data without the middleman.

Who can deny such power? Yet still relatively few use it routinely.

A recent TDWI Research report should convince skeptics and procrastinators that visualised data is as important as we’ve always suspected. In “Visual Reporting and Analysis,” by Wayne Eckerson and Mark Hammond and published in early 2011, visualisation won strong support from survey respondents: Sixty-seven percent rated its ability to improve user productivity as “high” or “very high.” Even more, 74 percent, rated its ability to improve business insight as “high” or “very high.”

One source quoted in the report said that visualisation makes it “much easier for the eye and the brain to understand what’s happening in the data. … It helps people see what is going to happen rather than what has already happened, and I think that’s a natural progression of analytics.”

A second source reports that users ask upon seeing visualised data for the first time, “Why didn’t we have this data before?”, when in fact they did have it before, but just didn’t see it. They didn’t even see it? They must have been using a spreadsheet. Much of that data never makes it all the way “between the ears,” says Ben Shneiderman, University of Maryland professor of computer science and longtime visualisation leader.

Machines can spew all the data they want to, but until it finds a home inside a brain, little happens. Once it does happen, writes Elizabeth Churchill, principal research scientist at Yahoo Labs, data analysis becomes “fun and game-like.” One question leads to a new, unexpected question. You go “down the rabbit hole,” she writes, “until you end up seeing your data world in a fresh way.”

Shneiderman, reports Churchill, believes that well executed visualisation — appropriately rendered, deep, and nimbly responsive to users’ questions — offers three main competitive advantages over spreadsheets.

First, decision makers can make use of strategies of discovery that tolerate missing and uncertain data. Second, use of time-series charts, maps, and social networks analysis lets users handle a wider-than-traditional variety of problems. Third, tight integration of data discovery into workflows “amplifies individual creativity” and brings along “catalytic benefits” of social creativity.

In fact, these insights he had in 2009 or before seem to be unfolding now. Visualisation has been evolving away from the old, static report-centred designs to more inviting, even game-like interfaces. Advanced tools are now more often designed now with intelligent users in mind.

Data is made to feel more approachable and easier to explore. Nimble, in-memory architecture encourages users to discover data in cascading questions while it frees IT from having to anticipate users’ questions and pre-aggregate data.

Once users found themselves untethered to IT, things really took off. With in-memory analytics, users can move among the data asking questions as they arise — allowing a kind of flow that encourages thinking, creativity, and collaboration. Decisions come faster, but better, too.

Today visualisation’s popularity is growing fast, Eckerson and Hammond observe, though it’s still “in the early stages of industry adoption.” Their survey finds that business intelligence users view data graphically just 35 percent of the time. That’s just over half the rate they view it textually, 65 percent of the time.

Even within organisations with visualisation tools in place, usually only a small number of people use them. However, report sponsors observed that demand is growing as business intelligence shifts “from static to interactive to collaborative.”

With such benefits, you would think the business crowds would beat down the door for visualisation. Why isn’t growth higher?

Reasons given in the report are simple. We could have guessed the first one listed: Old habits die hard. Many executives, managers, and business analysts have always used text-based budgets and plans created in Excel. They want to stick with what they know.

Such preferences may even go back much, much further than their familiarity with spreadsheets. Didn’t we feel we had grown up a little bit back in secondary school when we started using text-only readers? That’s a milestone we never forget.

Resistance may even make sense. One BI project manager quoted in Eckerson and Hammond’s report replaced numbers with bar charts — only to have sales people point out that they could see the same data just as easily the old way.

Then there’s the reason with roots in visualisation itself. Perhaps one of the least appreciated reasons for resistance is that some visualisations actually make analysis harder. Bad design or visual overload, cited in the report, can do that. Anything more than basic visualisation requires more skill than preparing rows and columns.

People tend to accept each other’s spreadsheet layouts. But when it comes to visualised data, we’re more demanding — perhaps because we really do see visualised data better. Visualisation expert Stephen Few devotes his first book, Show Me the Numbers (2004; Analytics Press), to the effective rendering of tables and charts. “They are so commonplace,” he writes in the introduction, “many of us assume that the knowledge of their effective use is common as well. I assure you, it is not.”

Ah, nothing is ever as easy as it looks. But while the current adopters learn how to visualise data, let’s let skeptics and procrastinators stick with spreadsheets. Sooner or later, natural selection will have its day.

Lou Jordano is TIBCO Spotfire’s director of product marketing. Lou directs Spotfire's globally distributed Product Marketing function to design, develop, and deliver its go-to-market strategy, messaging, positioning, and field enablement programs.