I used a gauge on a dashboard--and lived!

I hate to admit it, but I really am a data visualization purist. When I joined my current company, one of the first things I did was to attend an eye-opening seminar conducted by Edward Tufte. It changed my life.

Since then, I've read countless other sources on proper data visualization from a long list of authors. Sure, sometimes I think some of these heroes sound like Statler and Waldorf in their critiques, but there's no denying that our industry's awareness of the psychology of information consumer has had many positive effects.

This personal history made it all the more shocking to me when I violated one of my cardinal rules about dashboards--never use gauges! Gauges are the epitome of skeuomorphic design for a data visualizer. If the product is "a dashboard", then of course the user should feel like the pilot of a Boeing 747, right? Wrong!

When I think of dashboards with gauges, I usually imagine something that looks like this real-world example:
Dashboard with Gauges

While some might love this kind of thing, I could never present this kind of design to a client as a "well designed data visualization". It violates so many best practices, I just don't know where to start!

So how did I arrive at using a gauge?

Well, I was trying to find a good way to pack a lot of information in a tight spot. Since the target device was an iPhone screen, this is a challenge not just in space, but also in terms of navigation. The information I needed to pack in to less than 20% of a 4" iPhone display:

  • The cumulative values of a metric on four time scales (WTD, MTD, QTD, YTD)
  • The target for each time scale
  • Percentage to a target value
  • Visual indication of progress to goal
  • Visually differentiated (e.g. color/shape) indication when a goal was exceeded
  • Show both percentage and actuals
  • Provide navigational elements to "show details" for any of the four time scales

Wow! A lot to pack in to a small space, made more challenging that the UI is touch, so any navigation had to be large enough to be hit by a finger.

After running through all of the obvious best practices: bullet charts, sparklines, bar charts and the like, it occurred to me that a gauge control (the present but never before touched element in my toolbox) might actually work--if I could take a lot of the "gauginess" out of it.

So the following is the design I arrived at, which has the following benefits:

  • Packs a lot of information in a small space
  • Provides text, color and geometry that tells the data's story
  • Allows the user to quickly and easily ascertain which data is most important to examine
  • Since each "gauge" doubles as a large, finger-friendly navigational button, no additional space needs to be provided for drill-down switching

I'm still am not a fan of skeumorphic gauges--the kind with dials, radial tick marks and the like. But I realize I need to have a more open mind, and as our target devices proliferate, perhaps we need to revisit some of the data visualization tenets we hold as sacred.

iPhone Dashboard

BI visualization tools: interest trends over time

It's interesting to look at the trend of interest in Business Intelligence visualization/exploration tools over time. It seems that lately everyone is keen to talk about Tableau, and since their IPO interest in the product has seemed to grow exponentially in discussions I have with clients and colleagues in the industry.

Wondering if the broader infoweb bears out the same results that I see from within the industry, I did a quick Google Trends comparison of search interest of Tableau compared with the other end-user oriented tools that are most in conversations I have (QlikView, PowerPivot, Power BI and PerformancePoint).

Indeed, Tableau has been the top topic lately. Is this just awareness, or is there a major shift in mainstream BI preferences underway?

BI Front-end interest over time

To check out this visual yourself, click on this link to Google trends.

Office for iPad launches The Nadella Era

Yesterday Microsoft launched the long-awaited Office for iPad at a press event keynoted by Satya Nadella (read his remarks). I'm probably not the only one unable to remember the last time a Microsoft CEO tweeted his excitement about the adoption of software that runs on a non-Microsoft operating system!

Satya Tweets on iPad Office

Market Reaction (and Confusion)

There have been varied reactions to the product launch itself, ranging from elation to disappointment and anger (as observed by Paul Thurrotte). Some pundits even suggest that Office for iPad signals Microsoft's acknowledgment that its own tablet efforts have failed. I think that last one is unlikely, but everyone is entitled to their own opinion.

We do live in changing times. Microsoft under Bill Gates and Steve Ballmer was a company whose core strategy at times seemed to be the maintenance of an oligopoly made of Windows and Office, surrounded by hopeful investments--and a few hobbies.

It's no surprise that, during those two eras, Microsoft rarely created an important product that didn't seem partly designed to increase its core Windows/Office volume.

Even when it did ship products designed to support non-PC devices, those products typically worked very poorly with anything other than Microsoft's own Windows desktops. Readers who used certain versions of SharePoint know what I'm talking about! Many wonder if Microsoft is really committed to creating great software for competing mobile devices.

Today really wasn't about Office for iPad

Was Office for iPad to be shipped under Mr. Ballmer but simply wasn't ready before he left? Is it merely coincidence that Mr. Nadella was CEO when it was time to make this product announcement? Or, as many speculate, is Office for iPad another in a series of signals that a shift in strategy is underfoot at Microsoft?

To most, it seems more than mere coincidence. Today the new CEO released a product that many believe has existed for some time--and he did so enthusiastically. Not so long ago, we saw this new CEO make significant changes in his top ranks, and we know Windows Azure will soon be renamed Microsoft Azure, perhaps providing the most clear signal yet that the success of Microsoft is, as it should be, more important than the dominance of Windows.

I think we should not be surprised if CEO Nadella has a much broader plan for Microsoft than merely enhancing the value proposition for the Windows operating system. A new CEO's past predicts how he or she will shape the company in the future. This CEO, in his last job, was credited with removing the ties that required customers to purchase Windows along with its server, database and developer tools, indeed making them available as cloud services by subscription. In the same period, the Azure team created its Mobile Services BaaS offering which serves iOS and Android devices as first-class citizens next to its own Windows devices (a product I use to support cross-platform development).

But Microsoft is shipping many excellent software products for other products now. OneNote and Lync, for example, are exemplary applications on the iOS platform. It should be no surprise if Microsoft under Nadella focuses more of its energy on defining its relevance in a future software market rather than trying to extend the same dominance it had in the past. As a tech enthusiast and Microsoft shareholder, I couldn't be more excited about what lies ahead.