Warning: Highly Addictive Analytics - Part 2 - Economics of Intuition
Preface: This is a multi-part series on my transformation to a self-serve BI tool enthusiast/nut. I will post pictures and analysis in following posts. Read my other work here: www.ovum.com/authors/surya-mukherjee (some are behind the firewall)
Hi. My name is Surya Mukherjee and I am an addict.....of self-serve BI solutions. In this post, I will try to demystify the intuitive aspects of a solution. What is intuitive?
Intuitive is the new black
An intuitive user interface (UI) is one of the most important aspects of a self-service BI solution. The interface constitutes the proverbial ‘last mile’ of analysis, but for users, it is the first point of contact with the solution and the primary contributor to overall user experience (UX). Therefore, it is critical that self-service solutions invest considerable efforts in making their UI amenable to their target audience. Unlike a marathon, where the last mile ‘feels’ the longest, the target of self-service BI solutions should be reduce the amount of effort and learning a user needs to undertake to run the last mile.
But what is this intuitive you so speak of?
When you ask anyone how they know something is intuitive, the discussion takes an entirely new and sometimes ugly turn. Why is a Mac more intuitive than a PC? It just is! Well, I usually like a definitive answer. If you cannot measure, how can you improve? How can you tell a distributed development team them if they are on the right path? Apple enthusiasts typically start looking at me like I am an alien (and not the cute ET kind but the Independence Day kind) at this point. They foam and fume but fail to come up with a definitive set of guidelines to assess how intuitive something really is. I've read several publications, reports, and industry journals on intuitive design, yet, it is something very few venture to define.
The caveat is, of course, what is intuitive for me may not be the same for you. A developer's intuitive nirvana will be a novice's UX hell, and vice-versa. However, we are not talking about a non-homogeneous audience when we think of the targets for self-service BI users. We're talking of business users, who know a little or no Excel, understand the data they want to crunch, are frustrated with enterprise IT, and have a fair idea of how things work in the consumer world (most of them are super-users of Facebook or Twitter). Given so many common characteristics, is it really impossible to define intuitive for such an audience?
My and Ovum’s definition of intuitive BI
Building on several published interpretations of intuitiveness and our ongoing conversations with enterprises, I believe that a self-serve BI tool is intuitive when it has an appropriate combination of design attributes listed in the figure below.
What is the intuition quotient of your solution?
Source: Ovum
This should be fairly self-explanatory, but I will go into more detail in the next part of my blog, with actual examples from vendor solutions. At this point, my desktop is cluttered with installs from Qlik, Tableau, Spotfore, and more.
Clarity: The UI has standard elements that indicate what they are going to do.
Predictability: The UI is predictable, i.e. pushing a button which looks like a funnel will result in a data filter and not something entirely different.
Automation: The UI helps users to automate unnecessary activity and achieve common goals with the least effort.
Responsiveness: It is clear that the casual BI user will not wait more than a few minutes (at the most) for any result to be tabulated and more likely only seconds.
Defensive design: Another way of saying resilient software, or in other words, idiot-proof.
Familiarity: Well, at least a spreatsheet view, a tabular view, and so on - things that you are used to while analyzing data on something like excel.
Part 3 of this series will have screenshots from actual vendor solutions explaining who does what and how. Stay tuned.