BI for the Right Side of the Brain

We base Business Intelligence today on a clear premise: people are logical decision-makers. In this light, common BI practices promise that the right data, in the right format, at the right time, will help you toward better outcomes.

I see a problem with this view: as human beings, we just don’t work that way. What influences your knowledge workers’ outcomes? It’s not just the right data; it’s also the surroundings they work in, and their feelings as they decide.

You may be thinking we can’t do much to change the mind-set of our decision-makers. All we can do is provide that best data in the best format and make sure our visualizations and results are as precise as possible. Can we really control anything else? In fact, I do believe we can help to give users a more complete understanding of the data behind their choices.

First, however, let’s consider a couple of recent studies illustrating the effect emotions can have on decision-making. One Israeli research group, studying real cases, found judges granted parole in about 65% of cases after a meal break, dropping to as low as 0% just before a break or at the end of the day. The results were stark and clear. In a US study, judges were asked to set bail on some made-up cases. When a preceding questionnaire included a question about death, the judges set bail eight times higher than others who had not been asked anything on that topic.

I find such results quite scary! After all, judges are trained and experienced in making unbiased, regular, evidence-based decisions, are they not? What hope for the sales director, struggling with tactical choices in her quarterly review meeting?

In the world of decision support software, how can we support better decisions?

Knowledge workers already know one handy and familiar model: the simple, but potent, interface of the internet search engine. However, the use of search can seem counter-intuitive for a BI practitioner, because queries often return data point by the thousand and we know from experience how many prove less than useful. Yet, even though they deliver a near-overwhelming flood of results, search engines have a great advantage for the new and advanced user alike. They allow for browsing, or what Peter Pirolli has called “Information Foraging.”

Pirolli’s theory sees us foraging for information in a very similar way to animals hunting for food. We receive a great deal from the information environment, browsing through cues, or “information scent,” for the pieces we want. We may return more data with less structure – but enhanced by good browsing tools giving cues to content and relevance – and still provide better insights than presenting a small, focused, result set. Hunting and foraging represent basic human skills that help us to navigate the data surrounding us. You can get better results and better information through browsing and foraging compared to our accustomed structured queries.

We should also appreciate that the touch interface is a significant improvement to browsing. There’s a reason why orangutans in Milwaukee and Atlanta can use “Apps for Apes” and babies can use iPads. Touch interfaces feel satisfyingly intuitive. Everyone can understand the simplicity: using a fingertip to navigate and act. When applied to business, the possibilities seem endless. Learning from my own work in the field with customers, I know they happily spend more time navigating and exploring in a program when using a touch application, compared to a conventional desktop interface. We often call the touch interface of the tablet a relaxed “lean-back experience” compared to the more directed “lean forward” world of the desktop. More time spent foraging the data means more discoveries and insights, precisely because users are able to casually browse in a more natural setting. Furthermore, the touch experience is inherently exploratory. With a new tablet app, the first thing you are likely to try is exploring the interface and seeing what artifacts onscreen respond to touch, and which features reveal new information.

One retail mortgage company in the Netherlands deployed an app enabling customers to browse mortgage and insurance rates in the store. When the company switched from laptops to iPads, the conversion rate (consumers who decided to buy a mortgage in the store) increased by 35 percent. Customers enjoyed browsing and weighing their choices; perhaps as a result their final outcome carried greater confidence.

I believe Business Intelligence as a practice should not first and foremost aim to narrow the data space to a restricted set of predetermined answers, delivered in response to a well-defined query. If you can formulate that query well enough, then probably you have mostly found your way to the answer already. The exploratory foraging experience, what I often call the discovery experience, enables users to find things they didn’t even know they were looking for. Browsing opens up a world of information. As BI practitioners, that’s what we need to strive for. It’s not just about the data – it’s about the truly important insights we can uncover when our users are enabled with tools that support and expand our instincts.

Donald Farmer is the QlikView Product Advocate, working with customers and partners to establish QlikView as the leading solution for Business Discovery. Donald has over twenty years experience in analytics and data management. In that time he has worked as a consultant, in startups and as a leader of Microsoft’s BI product teams. He is a speaker at many international events on business intelligence, data integration and data management, blogger, and author of several books. In addition to his career in Business Intelligence, Donald has worked in fields as diverse as fish-farming and archaeology in Scotland. He is also a guest professor at SouthWestern University in Chongqing, China, and advises on several academic boards.