Making BI Analytics Fun
- By Stephen Swoyer
- April 17, 2012
The greatest trick Tom Sawyer ever pulled was when he convinced his friends to help him paint Aunt Polly's fence on a beautiful Saturday morning.
It was a trick, to be sure, but none of Tom's friends seemed to know (or even to care) that he'd been tricked. Painting a fence is hard work -- on a splendid Saturday morning -- with a 12-year-old's mindset -- is thankless work. The trick, of course, is that Tom made the painting of the fence seem like fun -- like a game.
There's a movement afoot to bring a similar kind of experience to business software -- called "gamification" -- and proponents claim that it has particular resonance in business intelligence (BI) and analytics, where the search for -- and discovery of -- insights already has a game-like feel to it. Gamification advocates want to amplify this effect: to intelligently apply game-like concepts and methods to BI and analytics.
This isn't to turn BI into "Angry Birds," stresses Donald Farmer, product advocate with QlikView. "If you can make software more like a game, you can encourage more immersion [in the software experience] -- you can encourage people to use it [the application] more," he argues, "so simply adding scores or leader boards doesn't do it. It doesn't get you immersion.
"Games aren't about scores: they're about the game play itself. You can't give users a [crummy] application with some scores at the end and expect that to work."
The (Game) Play's the Thing
In Twain's tale, Tom recruits new painters -- new players -- just as his first recruits are beginning to grow tired of the "game play" experience. On Farmer's terms, this is because Tom has a poor application. That he's able to keep the game going and recruit new users is a testament both to his charisma and to the reader's willingness to indulge Mark Twain.
Another example of what's meant by gamification is suggested by David Lightman, hero protagonist of the 80's cult film War Games. In the movie, Lightman nominally wants to play a specific game -- "Global Thermonuclear War" -- and is willing to go to great (and as it happens, highly illegal) lengths to do so. But whether he realizes it or not, the game Lightman most wants to play isn't "Global Thermonuclear War:" it's the experience of cracking and hacking systems -- with the nominal purpose of playing games before they're released.
According to Farmer, the actual experience of playing "Global Thermonuclear War" is anti-climactic. Lightman is instead hooked on everything that goes into being able to play "Global Thermonuclear War" in the first place. That's the immersive aspect.
"It's a question of game play: of how we can make [interacting with] BI more engaging. For example, you want to get people into the flow where they're asking questions continuously, where they're following [an analysis] from one question to another. Where questions lead to insights, and vice versa," says Farmer. "You can incorporate some [explicit] game play concepts -- scores or leader boards or achievement [metrics] -- but the most important thing is to deliver the kind of immersive experience that pulls the user in. That's the essence of good game play."
Consider a real-world example of applied gamification. The textbook case -- one that involves complex problem-solving and analysis -- is Foldit, an online game in which users compete against one another to solve puzzles. Foldit puzzles aren't ordinary puzzles, however. They involve protein folding. You know: the stuff for which Ph.D.s tend to win Nobel Prizes. Appropriately enough, Foldit was created as a joint venture between the University of Washington's Department of Biochemistry and its Center for Game Science.
In January, Scientific American reported that Foldit users had crowd-sourced the redesign of a protein that UW researchers had originally cooked up in the lab.
Gamers were able to achieve an eighteen-fold increase over the activity of the original design. That's much better than what the scientists themselves had been able to achieve. According to the report, researchers decided to take the problem to Foldit once they'd determined that their own efforts had "plateaued."
What's striking is that Foldit was designed with precisely this scenario in mind. It was designed for gamers, with the expectation that the experience of game play might produce something more. To the degree that an academic or researcher is a gamer or has an interest in playing games, a gamer can compete in Foldit puzzle-solving but will be competing against other players who don't necessarily share the same pure interest in (for example) protein thermodynamics.
They do share a pure -- and far more general -- interest in good game play, however. That's the point: in the context of game play -- i.e., of challenge, competition, recognition, and (of course) enjoyment and pleasure -- non-specialist players were able to far outstrip the best efforts of specialist researchers. And that's the impetus behind gamification in a BI or analytic context.
Advocates such as Farmer aren't even touting Foldit-like benefits. "I think one thing that's extremely important is that we're getting away from [focusing] just on user interfacing and concentrating much more on the overall user experience. We need to focus on making this [the user experience] much more immersive," he comments.
This is nothing less than an ontological shift, says Farmer: it's changing the context in which a person experiences, interacts with, and knows the (business) world.
"We absolutely need to move people from this mindset that says 'I have this query that I need to run,'" he argues. He says the convergence of several trends helps to make this trend inevitable. As mobile devices become pervasive, immersion and gamification will become ever more common. At the same time, advanced analytic and visualization technologies will become increasingly commoditized. These, too, will be incorporated into the immersive application experience.
Furthermore, don't forget the cloud: be it in the context of hosting, delivering (via AppStore-like venues that QlikTech and other companies are even now prepping), or enabling the collaborative experience, cloud computing will play an important role.
"Today, BI apps are always built to answer specific questions, and that's the wrong approach. A better approach is to look at [the issue] in terms of 'I need to know what's happening, and in order to do that, I need to be able to [contextualize] all of these other things that are happening.' It's about providing an overall awareness of not just that single hypothesis [that's formulated in response to a specific question] but an immersion in the data experience.
Farmer cites the example of a QlikTech customer in Japan, which he says was able to use QlikView to determine -- to use his words -- "what was happening" with its business in the aftermath of last year's devastating tsunami.
"They started noticing that [the tsunami devastation] was going to have all of these other effects that went beyond their [core] business," he explains. "Japan is also a major supplier of very critical components and critical ingredients ... so they were able to contextualize [the event]: to visualize it not only from their own [business] perspective but from [the perspectives of] their suppliers [and] their partners."