December 12, 2011

Churchill Club Presents The Big Data Effect

Copyright © 2011  Cary J. Calderone

Is Big Data being over-hyped?  "I certainly hope not" was Ping Li's heartfelt response to moderator Michael Chui's question to the panel (bios below).  Li's firm,  Accel Partners, made a splash in the news recently by announcing the creation of a 100 million dollar fund for investments in Big Data.   The Churchill Club panel members each gave their own interpretations of the Big Data Effect. They emphasized that Big Data was not just about  the volume of data, but how it could be researched, extracted, and analyzed.

Mr. Li, commented that Big Data was a mixture of machine-generated, user-generated, and open data, but the Big Data effect had to do less with size and volume and more with getting value from new data, old data, and, the mixture of the two.  Li believes "we are in the early days of this transformation" and he wonders "what will be the next Seibel or Cognos of the Big Data world?"  Gil Elbaz emphasized the important issue was whether the data would be open and transparent and therefore, easily accessed and exploited.  An interesting explanation of the Big Data effect was offered by Luke Lonergan.  Lonergan mentioned a unique situation where a large retailer, that had not yet bought into the idea of Big Data discovered it had customers, who because of their iPhones and Androids, knew more about the products than their retail clerks.  So even though one would think the retail clerk is there to help the consumer, because the clerks were relying on an antiquated terminal and sku system, they were in reality, an impediment to the sale.  To me, this particular Big Data effect might best be captured by the title of John Hagel's book, "The Power of Pull."   How can a company not recognize and move towards the expectations of its better informed customers if it expects to stay in business?  Slow moving retailers will be pulled towards Big Data by the expectations and capabilities of their customers and competitors.

My next favorite quote from the panel was that "privacy is the third rail of Big Data."  The advanced mining capabilities if Big Data are relatively recent and yet they have already evolved to the point where the average smart phone user would be dumbfounded to learn what collectors and facilitators of meta-data,  already know about them.  The panel noted that at a recent government inquiry, even the government representatives were surprised by how much information is already mined and how powerful it can be when analytics and metrics are applied with location.  This is especially true at what Ping Li finds  the most interesting new area, the cross-section between mobile and Big Data.  He believes the smart phone is the best data-capture device.  Anand Rajaraman added that it is not just about Big Data captured, but rather "fast" data and this is exactly what smart phones capture so well.  Mobile data is fast data.  In other words, it matters not only that the data can be captured and analyzed, but that the answers or results are calculated quickly.  This provides for the optimal Big Data effect.  Rajaraman also appreciates the potential and challenges at the cross-section between public and private company data.  "Do you bring all the public inside, or, move your private data to the cloud and risk security concerns?"   Keith Collins added that it was the flow of data that may lead to the best discoveries and innovations.

It seems the entire panel agreed that one of the important new areas for the Big Data effect would be advancements in the healthcare industry.  "The stores of information from real patients is a gold mine."  The old model of testing medicines on groups of people, or, a doctor relying on the patient's questionnaire for history, are archaic when compared to mining of healthcare information of most everyone in the world.  In this instance, size (of the data base) does matter.   The ability to quickly "crunch" every possible variance in patient history may prove as an essential element to the innovators.   My own favorite argument in support of the proposition is a simple one and involves Watson, the IBM computer that has been able to continually trounce very smart people in a game of Jeopardy.  Very simply, wouldn't you like to have Watson analyze your health for one minute, in addition to your doctor?    As smart as your doctor may be, can they cross-check possible side effects of the 3 to 4 pills you are taking as well or as quickly as Watson could?  Unlikely.   I believe everyone in the audience could identify a similar situation with their own healthcare experience.

Although it was not mentioned by the panel, my own personal experience has dealt with diagnosing food allergies.  We allergy patients have always had to self-monitor and self-diagnose to help the process.  I would love for that information to be readily available and accessible to every doctor I see.  Moreover, I would even offer it to a general database on allergies and interactions so others may benefit from my experience and I, from theirs.  To this end, the panel made the point simple.  More and better information available to the patients will lead to more accurate understanding of the conditions, and lead the doctor to making a faster and more accurate prognosis and, finally, achieving a better outcome.

In a recent blog post Seth Godin gave an example of the problems that still exist in healthcare, which he labels as "pre-digital."  Think about his frustration and imagine what Big Data technology may change.  Godon's description of a visit to the Emergency Room:
Six people doing bureaucratic tasks and screening that are artifacts of a paper universe, all in the service of one doctor (and the need to get paid and not get sued). A 90-minute experience so we could see a doctor for ninety seconds.  (full blog story)
It is easy to see how Big Data technology can improve a visit to the ER.  The ultimate challenge according to the panel, is whether there is a way to do it responsibly?   If so, open access to Big Data and personal health information is likely to garner quantifiable life-extending and quality of life improvements.

In addition to healthcare, Keith Collins explained how the Big Data Effect could help with energy policies and a smart grid by providing predictive and optimization analysis.  Soon they will have the capability to analyze the grid on 15 minute intervals.  He likes those who follow the creation of new regulations and then determine how to turn that into a value proposition.  Lonergan added that the new tools, smart meters and smart grids, have provided us with new data and new opportunities around that data.

In conclusion, whether it is healthcare, energy, or just plain old business productivity and profits, the final takeaway from this terrific panel is that everyone involved with Big Data will have to be very careful to safely navigate the third rail.  But, if they do it responsibly, then the innovation and growth potential will be very Big, indeed.

The Big Data Effect
Keith Collins, Senior Vice President and Chief Technology Officer, SAS
Gil Elbaz, Founder and CEO, Factual
Ping Li, Partner, Accel Partners
Luke Lonergan, Chief Technology Officer, Vice President and Co-Founder, Greenplum, an EMC Company
Anand Rajaraman, Senior Vice President, Walmart Global E-Commerce & co-founder, @WalmartLabs
Michael Chui, Senior Fellow, McKinsey Global Institute

No comments: