Faith in Finding the Way through LOTS of Data

“Sergey’s Search”, in this month’s Wired (http://www.wired.com/magazine/2010/06/storyboard-sergey-parkinsons/) starts with the fact that Google co-founder Sergey Brin knows he is at a higher genetic disposition to develop Parkinson’s disease.  It’s helpful that his wife is a founder of 23andme, an online genetic testing firm.  And not surprising that given his genetic predisposition, massive fortune, Stanford computer science background and subsequent Google enterprise, that he would invest in advancing Parkinson’s research with a non-traditional, information-theory/computing orientation.

This orientation inverts traditional biomedical research on it’s head, from an emphasis on small, “purer” data sets to “tons of data, a deluge of information, and then wade in, searching for patterns or correlations”, sometimes referred to as an “exaflood” of data.  The article mentions that Jim Gray (Microsoft research and computer scientist) believes that the evolution away from the practice of proposing and then testing an hypothesis to looking for patterns in the data would revolutionize scientific research.  Andy Grove, also a Parkinson’s sufferer, has called for a “cultural revolution” in scientific research.

Beyond the direct implications for biotech research (which is significant in itself), it also made me think about research in support of strategy development.  How often have you felt that someone had an answer or scenario in mind and was just looking for confirmation as you entered the data gathering stage?  I can remember on more than one occasion working with my colleagues to suspend the need to fill in the white spaces with new product ideas and to yield to wading in the thick of the information and see what patterns emerge.  The best strategies, that are robust enough to stand up to a multifaceted review (market, competition, core competencies, technology, distribution, etc.) and yield sustainable competitive advantage, are often the result of just such data explorations.