The Scourge of Unnecessary Complexity
One of the mottos of my work is “eloquence through simplicity:” eloquence of communication through simplicity of design. Simple should not be confused with simplistic (overly simplified). Simplicity’s goal is to find the simplest way to represent something, stripping away all that isn’t essential and expressing what’s left in the clearest possible way. It is the happy medium between too much and too little.
While I professionally strive for simplicity in data visualization, I care about it in all aspects of life. Our world is overly complicated by unnecessary and poorly expressed information and choices, and the problem is getting worse in our so-called age of Big Data. Throughout history great thinkers have campaigned for simplicity. Steve Jobs was fond of quoting Leonardo da Vinci: “Simplicity is the ultimate sophistication.” Never has the need for such a campaign been greater than today.
A new book, Simple: Conquering the Crisis of Complexity, by Alan Siegal and Irene Etzkorn, lives up to its title by providing a simple overview of the need for simplicity, examples of simplifications that have already enriched our lives (e.g., the 1040EZ single-page tax form that the authors worked with the IRS to design), and suggestions for what we can all do to simplify the world. This is a wonderful book, filled with information that’s desperately needed.
In 1980, a typical credit card contract was about a page and a half long. Today, it is 31 pages. Have you ever actually read one? Do you know what you’ve agreed to? Or more directly related to the focus of this blog, you probably waste a great deal of time each day wading through endless choices in software to find the few things that you actually need. For example, how many different types of graphs do you need to build an effective dashboard? Rather than the libraries of a hundred graphs or more that you’ll find in most products, I advocate a library of eight that satisfies almost every case. Software vendors fill their products with more and more features, whether they’re useful or not. Why? Because those long lists of features appeal to buyers. The designer John Maeda, however, observes what I’ve also found to be true: “At the point of desire, you want more, but at the point of daily use, you want less.”
That incomprehensible 31-page credit card contract can be distilled down to a page of clear information. The authors of Simple have done it. We could be spending more of our lives actually living and less time navigating endless confusion.
Not everyone wants simplicity, however. Some organizations and people thrive on confusion and use it to take advantage of others. We shouldn’t put up with this, but we do because we don’t think it’s possible to fix. It’s not impossible, but it won’t be easy.
We can’t all become crusaders for simplicity, taking on the challenges of improving government, the legal system, banking, and healthcare, but we can all do our part to simplify our own lives and organizations. To find out why this matters and how it can be done, read Simple. “Any fool can make things bigger, more complex, and more violent. It takes a touch of genius—and a lot of courage—to move in the opposite direction.” (E. F. Shumacher)
Take care,
9 Comments on “The Scourge of Unnecessary Complexity”
Now that’s a concept I can get behind. And very timely. Complexity brings anxiety with it. This reminds me of the question I hear so often this time of year,
“What do you want for Christmas?”
I want less.
Customers who need to make real decisions never ask for the fancy charts – they want to be able to make good decisions fast so it’s almost always a bar or a line chart that does the trick. Simple, yet can be highly sophisticated with the proper analytics.
Would you mind sharing which 8 charts you’re referring to?
Dirk,
Keep in mind that I’m using the term “graph” to refer exclusively to visual displays of quantitative data. The eight graphs that comprise a fairly comprehensive library for dashboards are:
1) Bullet graph (my version of a gauge, for featuring a single measure)
2) Bar graph (both vertical and horizontal, including stacked bars)
3) Dot plot (primarily used an alternative to a bar graph when you want to narrow the quantitative scale such that it doesn’t include zero)
4) Line graph
5) Sparkline (Tufte’s abbreviated line graph for time series)
6) Box plot (for comparing distributions)
7) Scatter plot (for displaying relationships between two quantitative variables)
8) Map (for displaying values in spatially)
Assuming that these eight graphs can be flexibly formatted, you’d have all that you need for most dashboards.
Stephen, thanks for a great post again! Hmm, about the list of preferable graphs, I would like to include the “micro calendar object” that you utilized in the dashboard design competition 2012 (a win-loss micro graph designed to show days of absences and tardies… that could also be replaced with a “sparkbar” graph).
In addition, one could argue that alert symbols are quantitative data, showing either binary data (yes/no) or quantified levels of risk (severe/moderate/no alert).
Janne,
What you’re calling a “micro calendar object” is merely a bar graph designed to display a value of 1 for an absence and -1 for a tardy. Products no doubt refer to this by various names, but essentially it’s a bar graph. Alert symbols I classify as icons: simple objects that convey specific meanings. Although quantitative values usually trigger alerts, the alerts themselves represent qualitative information rather than quantitative, such as “this needs attention,” or as you suggested, “severe alert, moderate alert, or no alert.”
With the exception of the red dot to highlight something worth talking about, I don’t think I have a lot of time for icons. It seems to me that they give people the illusion that they’re thinking systemically about the data, but what they’re actually doing is providing a simplistic prompt: Red! Bad! Green! Ignore (or congratulate oneself for one’s manipulation of the metric).
I suspect the RAG crutch encourages individuals to reinforce simplistic mental models. Kaplan & Norton tried to get round this (I think…) with their strategy maps, but this is too clunky in practice.
There is a limited amount of real estate in a BI dashboard environment, and when you factor in the necessity of making business intelligence available in a mobile environment that encompasses a plethora of screen sizes for smart phones and tablets, the issue becomes even thornier. So, how do we give our employees a simple way to cut through Big Data, find, sort and find, sort and filter Business Intelligence to get to information that helps them do their jobs and contribute to the bottom line of the company? Can dashboard software provide sophisticated tools that are easy enough for the average user to customize so that they can get personal alerts, design intuitive reports to share, use forecasting and predictive analysis tools, key performance indicators, balanced scorecards, graphic tools, cube management and all necessary tools with programmers or analysts?
Information overload and prohibitive complexity occurs for many reasons, not the least of which is simple intuitive data display, comprehensive integration of data and the ability to manage the dashboard software and business intelligence tools to accommodate growth and changing requirements.
We must make it easy for execs, managers and employees cut through the layers and get, view, filter, analyze, report and share data. They should be able to drill down into the data to find out what’s happening and easily share alerts, and clear reports and data with others who need to take action. Otherwise, the BI solution is a useless addition to the enterprise technology environment.
G’day Steve. Reading this post made me realise one of the reasons I resonate with you. We both share a value called ‘minessence’ (source: AVI, which is a values inventory I have used over my career to identify my personal values and learn how to guide my choices more in alignment with what is important to me, see http://www.minessence.net).
Minessence is the value of translating complex or abstract ideas into simple applications that impact the lives of others. You excel at this through your work. Thanks for the ongoing inspiration.
This is the #1 problem I see with novice dashboard-builders. My mantra is create “more” simple views and have them talk to each other versus a fewer complicated views of the data which seems to be what spreadsheets and legacy tools have taught people is necessary.