A few weeks ago I mentioned in this blog that I would soon announce the 2012 Perceptual Edge Dashboard Design Competition. Today, the competition officially begins. This will be the most challenging event of this type to date resulting in the most esteemed award for dashboard design (in my not-so-humble opinion) since I judged a similar competition for the B-Eye-Network back in 2006. This competition will serve several purposes:

1. A showcase for the current state of expert dashboard design.
2. An opportunity for me to use the submissions to teach best practices by critiquing several of them on this website and in the second edition of the book Information Dashboard Design, which I am currently writing.
3. An opportunity to provide sample dashboard designs that could actually be used to improve the quality of education in schools, for this competition involves the design of a dashboard that could be used by teachers to monitor the performance of their students.

The winning dashboard will be featured in Information Dashboard Design, Second Edition, due out during the first half of 2013, and in an article in the Visual Business Intelligence Newsletter. No, you won’t win $10,000 or an all-expenses-paid vacation to the Bahamas. Instead, you will have an opportunity to advance the information age by showing a better way to display data for performance monitoring. In other words, you will have a chance to do something useful for the world.

Here are the basic facts:

• The deadline for submissions is September 28, 2012.
• Submissions should be emailed to info@perceptualedge.com.
• Multiple entries from an individual are allowed.
• Submissions will be judged by me—Stephen Few—with the assistance of two subject matter experts in student performance monitoring. The identities of the competitors will be hidden from me until after the winners have been selected.
• Anyone may enter the competition, including employees of vendors that sell dashboard software.
• You may use any software that you wish to design this dashboard. In fact, you are welcome to design a dashboard using drawing software if you’d like, such as Adobe Illustrator, which would free you from the constraints of a particular dashboard tool. The purpose of this competition is to assess your dashboard design skills, not the merits of a particular tool.
• The student performance data that will be used is provided in an Excel file, which you may download be clicking here. The file contains behavior, aptitude, and achievement information for students in a single high school mathematics class.
• By submitting a dashboard to the competition you are granting me the right to include it my courses and publications.
• The winner will be chosen based on one fundamental criterion: the degree to which the dashboard could be used by a teacher to rapidly and effectively monitor the performance of her students for the purpose of helping them improve their mathematics skills.

All other information that you’ll need to participate in the competition is included in the Excel file mentioned above.

I encourage you to enter the competition and to take it seriously. The world needs better ways of monitoring information, and you can help by applying your skills to this task. Whether you win the competition or not might matter to you, but it isn’t the only thing that matters. Either way, you will learn a great deal through the process.

Take care,

This week I’ve been relaxing in Montana. Two days of my visit here so far were dedicated to learning the basics of fly fishing. During my childhood, the only recreational activity that my family engaged in together was fishing — mostly at lakes using bait to catch catfish, bass, and perch. In recent years, I’ve wondered if fly fishing might reconnect me to the joy of fishing in my youth in a way that better suits the person that I’ve become: one who is nourished by intellectual challenges. Based on the recommendation of a friend from Montana, I engaged a fellow named Joel Thompson of Montana TroutAholics to serve as my guide and teacher in the ways of the tied fly. To my delight, Joel turned out to be more than an expert fisherman.

If I ever move to a place with an abundance of pristine rivers and streams like Montana, I suspect that I’ll take up fly fishing as a regular form of meditation. In the meantime, I’ll treasure the insights that Joel shared with me, especially those that, odd as it might seem, apply to my work in data visualization.

Joel has observed through many years of fishing that the flies that effectively appeal to trout, despite their endless variety, can be reduced to about 10 basic patterns. These patterns are limited in number because the senses and appetites of trout are finite. Flies mimic the appearances of insects that trout eat. The patterns that express the appearance of these insects (mayflies, grasshoppers, etc.) can be distilled to a few, but their expression can vary a little in specifics such as color.

I told Joel that my work in data visualization similarly revolved around a finite set of visual attributes that could be used to encode quantitative data. Just as trout have a limited set of visual patterns that appeal to their appetites, so humans have a limited set of visual attributes that can be used to represent data.

Joel went on to explain that flies come in endless variety in part because people tend to design them in ways that appeal to themselves, which considerably exceeds the interests of trout. People love to make them colorful, fanciful, and pretty. They overwork the design, adding more and more features that amount to pure decoration. He admitted that even he is naturally inclined to overdo it when designing a fly. When he visits a fly fishing shop or online fly store he’s automatically drawn to flies that look cool, but knows that the essential question is, do they look cool to trout? He has to rein himself in when wrapping his own flies — a pastime of many fly fisherman, especially during the cold and dark winter months — reducing his designs to clear expressions of the basic patterns.

While visiting a fly fishing shop in Missoulla, I noticed that the two clerks were occupied, giving every ounce of their attention to an attractive young female customer — apparently a rare cause for celebration in a the male-dominated world of fly fishing. After she departed, the guys enthusiastically shared their joy with me and another male customer by recounting the interaction. They told us that upon entering the store the young lady asked for assistance in selecting “a pretty fly.” In her naiveté she expressed what many a stoic male fisherman often feel but would never admit: a natural attraction to things that are pretty and cool.

Novice or avaricious makers of flies approach the task as the addition of features rather than the subtraction of all that goes beyond the essential pattern of mimicry. I suspect that expert fly makers appreciate the elegance of simplicity. The parallel to data visualization is obvious. Data visualization beginners and those with experience who are willing to do what appeals even if it doesn’t work, make eye-catching charts that attract parts of our brains that are not engaged in understanding data. We can forgive the naiveté of beginners if they’re open to learning. The avarice of vendors that make dysfunctional data visualization tools and of so-called data visualization experts who obscure and complicate simple data-based stories by dressing them up in bangles and glitter to earn the adoration of a naïve public, this we shouldn’t forgive.

Whether we’re trying to lure a trout to the end of our line or working to make better decisions based on data, we need to focus on the patterns that work. Overworking them might be a natural inclination, but so is over-eating. It’s time to behave like responsible adults.

Take care,

Forrester has released a new report titled “Advanced Data Visualization (ADV) Platforms” by Boris Evelson and Noel Yuhanna, which you may purchase for a mere $2,500. I suppose that some reports might actually be worth the price of 50 books on the topic (at $50 each), but I’m certain that this report isn’t on that list. How can I be certain without reading it? (No, I didn’t pony up $2,500.) Because, based on his previous work, I know that Boris Evelson understands little about data visualization and is misinformed in many respects, and because the abstract that Forrester’s made available to tempt us into buying it reveals that it’s fundamentally flawed. Here’s what it says:

Enterprises find advanced data visualization (ADV) platforms to be essential tools that enable them to monitor business, find patterns, and take action to avoid threats and snatch opportunities. In Forrester’s 29-criteria evaluation of ADV vendors, we found that Tableau Software, IBM, Information Builders, SAS, SAP, Tibco Software, and Oracle led the pack due to the breadth of their ADV business intelligence (BI) functionality offerings. Microsoft, MicroStrategy, Actuate, QlikTech, Panorama Software, SpagoBI, Jaspersoft, and Pentaho were close on the heels of the Leaders, also offering solid functionality to enable business users to effectively visualize and analyze their enterprise data.”
[emphasis mine]

No one with even a modicum of expertise in data visualization would place IBM, Information Builders, SAP, and Oracle in the list of advance data visualization leaders along with Tableau Software, Tibco Spotfire, and SAS, nor would they leave good vendors such as Panopticon and Advizor Solutions off the list. The chasm that separates these data visualization providers is huge. Lists and claims like these are typical in reports by Forrester and Gartner. There’s a reason for this. Do you know how software vendors get onto Gartner’s annual BI magic quadrant? They pay for the privilege. When vendors that deserve notice are missing from the list, it’s because they haven’t paid the fee. When they suddenly appear in the magic quadrant although they were never there before, it isn’t because they’ve improved in a significant way, but because they paid the fee for the first time. I don’t know exactly how Forrester works, but I suspect that it’s the same. (Anyone who works for Forrester is welcome to tell us differently by responding to this blog.)

Even though I haven’t read the report, I did get a glimpse into it by reading Evelson’s recent blog post about it. What does Forrester mean by advanced data visualization?

How is Advanced Data Visualization (ADV) is [sic] different from earlier generations of data visualizations? Many corporations have effectively used — and will continue to use — traditional business graphics, such as bar charts and pie charts. At the next level, modern technologies have enabled the use of more dynamic and interactive business graphics, such as real-time dashboards and charts that update automatically as the data changes…Now, through ADV, potential exists for nontraditional and more visually rich approaches, especially in regard to more complex (i.e., thousands of dimensions or attributes) or larger (i.e., billions of rows) data sets, to reveal insights not possible through conventional means. Forrester differentiates ADV from static graphs and charts along six capabilities, as follows:

1. Dynamic data content.
2. Visual querying.
3. Multiple-dimension, linked visualization.
4. Animated visualization.
5. Personalization.
6. Business-actionable alerts.

What becomes immediately clear to anyone with experience in data visualization is the fact that by “advanced” Forrester is referring to features that have existed for quite awhile. Tableau, Tibco Spotfire, SAS (the product JMP in particular), Panopticon, and Advizor Analyst have always supported these features, with the exception of animation, which is only useful for some visual storytelling, not for data exploration and analysis, and “business-actionable alerts” (don’t you just love the jargon?), which are useful for performance monitoring. The bar is set fairly low. Forrester is just talking about basic information visualization as defined by Card, Mackinlay, and Shneiderman in 1999 as “…the use of computer-supported interactive visual representations of abstract data to amplify cognition.”

But wait, there’s more.

Navigating the ADV landscape requires evaluating significantly more features than the six key ADV capabilities described in the previous section. In our latest research, Forrester identified numerous functional and technical capabilities businesses need to architect, design, build, and implement ADV applications. Forrester recommends starting your evaluation of ADV platforms by defining your requirements for the following functionality:

• Types of graphs, charts and other visualizations.
• Tufte’s microcharts.
• Cockpit gauges. [Seriously?]
• Visual query.
• Visual exploration. [Duh!]
• Geospatial representations. [This is just one of those useful forms of data visualization that belongs among the “type of graphs, charts and other visualizations” that they referred to in their first point above.]
• Modes of interaction.
• Storyboarding fit for client and boardroom-level presentations.
• Data latency.
• Data granularity based on your requirement.

The points in the list above — those that are useful — are well known and have been for ages. Granted, they are not well known to everyone in the business intelligence community, but should these folks really pay $2,500 for this report to learn about them when excellent books and many free resources are available, which were written by actual experts in the field?

Typical of many in the field of business intelligence, Forrester’s team is obsessed with the technology rather than the skills and activities that inform effective data visualization, especially of the advanced variety. This fact is revealed by the following section of Evelson’s blog:

Forrester identified eight categories of ADV technical architecture capabilities through posing the following questions:

• What analytical engines does the ADV platform support? How does it access and process data?
• Is there an intermediate storage platform?
• How is the in-memory data model managed?.
• What types of data can the ADV platform analyze?
• Does the ADV platform support write-backs?
• What platform/technology is the ADV output based on?
• What, if any, ADV-specific programming language is used?
• What are the ADV platform’s integration capabilities?

Most of these are the wrong questions. If either of the two authors of this report and any of the seven contributors had any real experience with data visualization or had ever worked as data analysts, they would realize this. Read the biographies of the contributors to this report and you will discover in each case someone who is entrenched in an old and myopic BI mindset. To understand the potential of data visualization, especially advanced data visualization, you must actually study it and do it. In many respects, an extensive background in BI, especially at an expert level, works against one’s ability to understand the interaction with data that is needed for sensemaking. It is a human activity that involves our eyes and brains. It can be supported wonderfully by well-designed technologies, but never driven by them. To date, none of the software vendors on Forrester’s list, except for the three that I singled out as worthy, have developed better than rudimentary data visualization tools, many of which are truly awful.

If you want to learn about data visualization, especially what to look for in a well-designed tool, read the article in my most recent Visual Business Intelligence Newsletter titled “Criteria for Evaluating Visual EDA Tools“. You will not only learn more than you would by reading Forrester’s new report, but you will also save yourself $2,500. Now that’s a deal. If you’ve already paid for Forrester’s report, demand your money back. Make them accountable for delivering a level of expertise that warrants the high price.

Take care,

From time to time someone characterizes my work with the phrase “less is more.” (Sometimes, those who enjoy wordplay incorporate my name into a variation of this expression — “Few is more.”) When I hear this, I always cringe a little. I have never used this expression to characterize my work or design philosophy. Even though the spirit of the phrase is in many ways consistent with my work, it is an oversimplification. While it is sometimes true that less is more, it is also sometimes false, depending on what is lessened. Less of something that is useless is definitely more. For example, less distraction in a data display produces greater perceptual efficiency and often more understanding as well. In fact, the better expression when referring to useless content or embellishments in charts is perhaps “none is more.” Less of anything that is useful, however, is never more when it’s incomplete.

This applies especially to complexity, which often exists in information. Complexity is neither good nor bad in and of itself. Necessary complexity — that which is meaningful and relevant to the task at hand — is useful and should therefore never be eliminated or even reduced. Instead, it should be managed. When presenting information, we manage complexity by finding the simplest possible way to display it, never crossing the threshold into over-simplification. This can often be done by breaking the information down into logical and meaningful parts and presenting each part separately at first. Once your audience is comfortable with the parts, then you can combine them, perhaps one at a time, to build up to the full level of complexity in a way that people can absorb without ever being overwhelmed. Sometimes we can manage the complexity of a wealth of information by the way we organize it. For example, several concepts and facts that relate to one another can be organized on the page or screen in a way that makes the nature of the relationships clear. This is what we’re attempting when we organize text into an outline to reveal the hierarchical relationships that exist between parts of the content.

The expression “less is more” fails because it often ignores useful complexity that exists in information. Less of what’s useful is not more. More information is only more when more is needed. Less information is only more when less is all that’s needed. Need I go on?

Take care,

Two distinct approaches to presenting data graphically exist today—data visualization and data art—and rarely do the twain meet. They differ in purpose and in design. When we fail to distinguish them from one another, we not only create confusion, but do great harm as well.

There are as many definitions of data visualization as there are definers, but at the root of this term that has been around for many years is the goal that data be visualized in a way that leads to understanding. Whatever else it does, it must inform. If we accept this as fundamental to the definition of data visualization, we can judge the merits of any example above all else on how clearly, thoroughly, and accurately it enlightens.

By data art, I’m referring to visualizations of data that seek primarily to entertain or produce an aesthetic experience. It is art that is based on data. As such, we can judge its merits as we do art in general.

Either one, done well, is worthwhile, assuming that it fits the task at hand. If the task is to help a particular group of people understand something, then data art is not appropriate, no matter how well it is executed. If the task is to entertain or engage an audience in a particular emotional experience, then data visualization probably isn’t appropriate. If the situation requires that both objectives are achieved, then a deeply informing and aesthetically beautiful visualization would be in order. Although it is quite easy to make any data visualization aesthetically pleasing, it takes a great deal of skill as a visual designer and information communicator to make one beautiful.

People make better decisions when they’re based on understanding. For information to be understood, it must often be presented in visual form. This is because patterns, trends, outliers, and a sense of the whole as opposed to its parts require a picture for the human brain to see and comprehend. Data visualization is essential. Visualizing data effectively is vital. Anything less is frivolous, costly, and harmful.

How in particular is data art—visualizations that strive to entertain or to create aesthetic experiences with little concern for informing—harmful when it masquerades as data visualization?

1. It suggests that data cannot be visualized without training in the graphic arts. As such, it works against the democratization of data. In fact, anyone of reasonable intelligence and a little training can present data effectively. It’s vital that this ability spreads more broadly across the population, because it can play a role in making a better world.
2. It features ineffective practices as exemplars of data visualization. It encourages people to present data in ways that are difficult to perceive and understand simply because they are prettier or more entertaining, which is rarely relevant to the task.
3. It keeps the practice of data visualization spinning its wheels, never able to progress beyond the mistakes of the past. Best practices of data visualization have emerged through many years of research and experience. “Those who cannot remember the past are condemned to repeat it” (Santayana).

I am personally and painfully acquainted with each of these problems. For this reason, I try to differentiate data art from data visualization and encourage others to do so as well.

Take care,