In-Depth
Q&A: Data Strategies and Philosophical Dilemmas
We examine how creating a data strategy has philosophical implications, raising ethical questions about analytics and emphasizing the need for data governance.
[Editor’s note: Frank will be presenting the opening keynote again at the TDWI World Conference in Las Vegas on February 13, 2012. This is the second time Frank will be a keynote speaker; his first presentation was in 2010.]
BI This Week: There was a bit of excitement following your keynote presentation at how the TDWI World Conference two years ago.
Frank Buytendijk: Yes, I had to leave in a bit of a rush, as our youngest daughter was about to be born. And my flight was delayed, so I missed my connection home! I was home just in time for the birth of Wilhelmine, who will soon be two years old now, so I am really happy to be back and see more of the conference this time.
Your presentation in 2010 was about strategy and dilemmas, and how a deep analytical approach sometimes can even be counterproductive. In reading your book, Dealing with Dilemmas afterwards, I noticed quite an emphasis on ethics. Is that the reason why you continued studying philosophy?
In a way, it is. Research is a continuous path of discovery. My third book, Performance Leadership, was about organizational behavior, and I ended up redefining the concept of an organization. We always assume that an organization is a group of people with a common goal. Have you ever seen a group of people with a common goal? Still, all of our information management practices are based on this notion. I think an organization is a unique collaboration of stakeholders that understand that to reach their goals they must by work together. The goal of the organization is to find ways reconcile the different goals of stakeholders. This leads to very different information management strategies.
This was the starting point of the fourth book, Dealing with Dilemmas, which looks at how organizations on the strategic level can reconcile different, often conflicting, goals between stakeholders. Shareholders want something different from what than employees want, customers have different requirements than suppliers, and so forth. Strategic decision-making typically affects multiple stakeholders, and this automatically means there is the question of morality and ethics, which are important philosophical themes.
I have no formal background in philosophy but I was fascinated immediately, so I decided my fifth book, tentatively titled In Search of Wisdom, should be about philosophy and technology, examining questions such as what is real, what is true, and what is good. It will be a collection of essays exploring how the old philosophers would have commented on modern themes in business and IT.
What does philosophy have to do with business and technology?
Everything! More than a third of the top 100 largest economic entities in the world are not countries, but enterprises. Businesses simply cannot be amoral; they affect society deeply. Technology has permeated our lives so deeply that we cannot imagine living without it anymore. One way of defining the purpose of technology is that it should augment human capabilities. Yet, many of the business systems I know just make you dumb. Philosophy should be fun particularly for IT professionals; they have a lot in common with philosophers. IT professionals like to think their way through problems, and they are often focused on an ideal world, how things should be. IT professionals also like to think conceptually. These are all things that many philosophers do as well.
How are these views related to having a data strategy?
The more you think about it, the more creating a data strategy has philosophical implications. For instance, I have been arguing that we live in the days of postmodern IT, where there is no place anymore for a single version of the truth. Postmodernists claim there is only perception, and truth is nothing more than a sufficiently shared perception. With Big Data being a reality, and with data sets becoming increasingly complex and distributed, even the computer’s response to a query becomes nothing more than an opinion. We can’t really track the background of the results anymore, and the same query to a different but similar data set will give a different opinion.
Creating a data strategy nowadays starts with realizing the postmodern reality. We should move away from the idea that “the numbers speak for themselves” and create a data strategy based on collaboration, triangulation, and discussion -- all the things that were part of the 2.0 hype but never had a real business case. Furthermore, I hope that data strategies will start to include the ethical ramifications of data analytics.
The ethical aspects of analytics sounds interesting. Can you share best practices?
That is the weird thing -- there aren’t any. I know of one large project in the pharmaceutical industry where they use extremely advanced analytics on clinical trial data. Suppose you do the analysis on the granular level and you identify some correlations between socio-demographics, DNA data, and certain medical conditions. Sharing that information with patients has ethical consequences. How much can you tell people about their specific (yet statistical) future? Not sharing that information has ethical consequences, too. Analyzing the data on a higher aggregate level has the same ethical issues, while you could have. There are no clear-cut answers. This may seem like a pretty far-fetched example, but in my keynote and in my workshops I will share many examples of ethical consequences of analytics that concern us all.
It seems you are emphasizing the need for data governance -- what data can and cannot be used for.
Absolutely! We all talk about data governance but hardly anything has been taken care of -- in business, but even more in public sector. Government can be seen as a “social contract,” the agreement that we as citizens have with each other. This contract takes care of two things: there are some rules and laws we should obey, and we pay taxes.
One of the main tasks of government is to provide freedom and safety for all. There is a continuous tension between those two, and the balance -- if there is any -- is shifting towards less freedom.
Governments invest heavily in data integration and analytics, but what if it goes wrong? We haven’t even seen the beginning yet of the consequences of identity theft. Furthermore, we need to answer some tough questions about data ownership: Who owns our medical data? Who owns our financial data? Who owns location data? Mostly this will be an educational process by making costly mistakes, public outrage, and even companies going bust, breaking unwritten rules and crossing unknown lines.
What can we expect from your keynote and workshops?
The keynote is going to be a fun start of the week. Lots of stories, lots of examples to unfreeze the mind and make you think -- the perfect start for a conference. I believe we should treat the topic seriously, but why not with a smile? My workshops will be mostly a discussion and work on cases, very active and participative. However, philosophy is dangerous. Once you dive into that world, there is no turning back. Descartes warned us already: you’ll start doubting everything, except perhaps doubt itself. Not for the faint at heart.
Frank Buytendijk’s research focuses on strategy, organizational behavior and performance management. With his background at Gartner (where has was a Research VP) and Oracle (where he founded and ran the global thought leadership program), he is a well-known authority. His research and presentation style is often described as out-of-the-box, slightly provocative and entertaining. Follow Frank on Twitter (@FrankBuytendijk).