Image courtesy of conference organizers Richard Wang PhD and Gjertrud Djupvik.
Larger than Big Data
“Evolving Data for Organizational Performance” was the theme and diverse case studies were discussed. At GlaxoSmithKline (GSK), former chief data officer Mark Ramsey spoke of the difficulties of using their vast quantities of data––including more than 2,000 structured data sources, 1,000 unstructured data sources, and more than 1 billion files—using outdated data management procedures. He found that both extract, transform, load (ETL) and master data management approaches were too laborious to produce actionable intelligence. A new framework using automated discovery and pipeline orchestration produced a better outcome. Results include the standardization of 1,000 clinical trials for cross-study analysis, a single integrated data platform for biological and chemical data, and 17 new deployed solutions for business operations.
Yet corporate success wasn’t the only star of the show. The first Chief Data Officer at the Department of Defense, Michael Colin, told both literal and figurative war stories on his work to improve readiness while increasing efficiency at the US Department of Defense. With an IT Budget of $46.3 billion and thousands of decades-old data centers, he half-jokingly referred to his job as “an archaeologist in the dig” dealing with a “but wait, there’s a harder way” operational approach. Despite these challenges, Colin has achieved measurable improvements in controlling costs and increasing aircraft availability. A large part of his success stems not from using fancy new tools but from cultural shifts and data literacy improvements.
The People Factor
The human part of the data equation was also stressed in a presentation from Jo Ann Stonier, Chief Data Officer at Mastercard. With an interesting background in not just IT matters but also interior design, she spoke on the need to balance innovation with caring for the needs of individuals and society. Her experience with the design thinking framework of keeping development iterative and flexible based on real world behavior has led to increased data literacy in the company and demand for more resources.
Stonier also addressed the company’s commitment to social concerns through a program called Data for Social Good, which is based on the tenant that information inequality is one of the major social issues of this century. In a public-private partnership, Mastercard supplies data science expertise to bridge data suppliers, such as companies, governments, and universities; with charities, NGOs, humanitarian aid groups, and researchers with the goal of solving social problems.
It's the Little Things
Tom Davenport is the closest thing to a rock star in this industry, with serious credentials to boot (President’s Distinguished Professor of Information Technology and Management at Babson College, numerous articles in Harvard Business Review, and Fellow at the MIT Initiative on the Digital Economy). In his “The State of Enterprise AI: Think Big, Start Small, Scale Up, Skill Out” presentation he stressed that collections of “low hanging fruit” are prevailing over “moon shots”.
He cited several examples of compelling projects that didn’t quite make it along with smaller scale projects that did within the same organization, such as the MD Anderson Cancer Center using Watson to treat certain forms of cancer. $62 million dollars was spent without electronic health records integration—the project has been put on indefinite hold. Yet they were successful using “cognitive scale” technology to identify patients who needed help paying bills, increase operational efficiency, and improve customer satisfaction.
Resources for a Role in Flux
A recurring theme at the event is that while demand is skyrocketing for Chief Data Officers there is no clear-cut definition of what the role does or whom those holding the title report to. Some focus solely on data governance issues while others have broader responsibilities including data procurement, risk management, and analytics. Some report to the CEO while others report to the Chief Information Officer or Chief Digital Officer (although this title is diminishing in use). In an effort to bring a common body of knowledge to the field, conference chairman Dr. Richard Wang began the Chief Data Officer Institute to explain the definition and landscape of the CDO role and provide introductions to big data technologies, data governance, analytics, and data integration. The various courses balance strategic concerns with hands-on learning to prepare new candidates for challenging careers.
The MITCDOIQ will be back for its 14th year in July 2020, this time at a larger venue to prevent another sellout. In the meantime, you can read more on their website or follow the organizers at @MIT_CDOIQ on Twitter.