Why managing data scientists is different — Roger Stein

MIT Sloan Senior Lecturer Roger Stein

MIT Sloan Senior Lecturer Roger Stein

From the MIT Sloan Management Review

While businesses are hiring more data scientists than ever, many struggle to realize the full organizational and financial benefits from investing in data analytics. This is forcing some managers to think carefully about how units with analytics talents are structured and managed.

How can organizations realize the promise of the evolving disciplines that we broadly call analytics?

Although financial firms were among the first to recruit “quants” to use sophisticated mathematical models and high-powered computing hardware, analytics groups have now taken hold in areas ranging from health care to political campaigns to retailing to sports. Organizations like these can benefit from the insights gained by financial service firms on how best to manage teams doing advanced analytics. It requires skills and philosophies that are different from those that arise in managing other groups of smart professionals.

Rather than just involving oversight and planning, managing a data science research effort tends to be a dynamic and self-correcting process; it is difficult to plan precisely either a project’s timing or final outcomes. For those unused to this type of work, this process can seem quite messy — an unexpected contrast to a field that, from the outside, seems to epitomize the rule of reason and the preeminence of data.

Read More »

How bad data fed the Ebola epidemic — Rachel Glennerster, Herbert M’cleod and Tavneet Suri

MIT Sloan Associate Prof. Tavneet Suri

MIT Sloan Associate Prof. Tavneet Suri

From The New York Times

The West African Ebola outbreak first hit Sierra Leone in May 2014, followed by an explosion of cases in the capital Freetown in the autumn. The epidemic now counts more than 10,500 cases across Sierra Leone, with signs that the spread is slowing.

The early days of the crisis were characterized by a sense of immense fear, anxiety and alarm, regionally and globally. In Sierra Leone, a three-day, countrywide, military-led lockdown in September fed the fear in West Africa and beyond. Many flights originating in unaffected African countries were restricted. African students were prevented from attending some American schools, and there were countless reports of discrimination against Africans across the globe. Pictures of health workers in full protective suits became a ubiquitous symbol of the panic.

Misleading reports, speculation and poor projections from international agencies, government ministries and the media about the Ebola outbreak exacerbated the problem. The fear that was spread by the dramatic reports that accentuated the negative, undermined confidence, made it harder to encourage people to seek care, and misdirected attention away from Sierra Leone’s urban areas, where data suggest the economic effects of Ebola have been concentrated.

Valid, credible and timely data is essential during a global crisis. Without reliable data, efforts to assist affected people and to rebuild damaged communities can be misdirected and inefficient.

Read More »

New MIT Sloan Management Review study: An advanced analytics culture outweighs all other factors — David Kiron

The Need for Culture

The Need for Culture

What distinguishes the winners from the losers among companies converting data and analytics into a positive force in their strategies and operations? And what practices are keeping the winners ahead?

The Analytics Mandate, a new research report from MIT Sloan Management Review and SAS Institute, takes several steps toward answering these questions.

Our most significant finding? Our study shows that an advanced analytics culture outweighs other analytics-related factors -including data management technologies and skills-among companies that strongly agree they are gaining a competitive advantage from analytics. Essentially, a strong analytics culture is the lynchpin in moving from competitive parity to competitive advantage.

The need for change within a corporation’s culture, and the best way to achieve it, are both nicely illustrated in a case study included in our report.  WellPoint, the largest for-profit managed care organization within the Blue Cross Blue Shield umbrella, knew that sharing insurance data with physicians would provide doctors with a 360-degree medical view of every patient. This in turn, would better enable them to spot patients more likely to go to the emergency room or be readmitted to a hospital, contributing to expenses that drive up the high cost of health care delivery.

Within WellPoint, creating the data reports for physicians initially became a classic showdown between IT and interests from the business side.

The initial reports, prepared by the IT team, were late and lacked fundamental functionality.  For instance, different units within the company reported an emergency room visit in different ways.  The IT team’s explanation: no one told them the definitions had to be the same. This much was true — the business side didn’t think it needed to specify that emergency room visits be consistent across reports. They had assumed this was a given.

The high-profile project was subsequently placed in Red status. At this point, senior management got involved. Problems were brought to executives who, in turn, ensured resources were allocated. Outside consultants and experts were hired. More resources were diverted to the project.

Finally, after many challenging discussions, IT and the business side began working together using an iterative development approach called “Agile”, which focuses on “user stories.”. This meant understanding the perspective of the end user—the provider—and the context in which he or she would be using the data, as opposed to just developing according to a static set of  requirements.

Early reaction to the data system from doctors has been highly positive.  Over time, WellPoint believes that the proactive, coordinated-care model made possible when providers have actionable insights at their fingertips can cut health care costs by as much as 20%. That could work out to billions of dollars, given that WellPoint reimbursed more than $99 billion in health benefits for commercial and individual members in 2013.

In short, to create strategic benefits with analytics WellPoint had to change its organizational behavior. Without an effective collaboration between the business side and IT, the program would have remained in jeopardy. Without leadership’s involvement, the program would have remained in jeopardy. Preparing data for a strategic role often means changing business conduct and that, more often than not, requires a top down process to create the necessary alignment of incentives and goals.

To read the full report, please visit “The Analytics Mandate.”

David Kiron is Executive Editor, Big Ideas initiatives, for MIT Sloan Management Review.

Maybe the search for the Malaysian Airlines plane needed a chief data officer — Stuart Madnick

MIT Sloan Professor Stuart Madnick

MIT Sloan Professor Stuart Madnick

From Quartz

The search for an airplane lost on a 2,500-mile international journey requires consolidating information from many organizations, both public and private, from all over the world. It involves analyzing vast amounts of radar, sonar, and satellite data, coming from many diverse sources, including military bases, air traffic controllers, naval ships, and other airplanes.

What if the authorities investigating the missing plane had been prepared to manage big data the way many corporations do? What if the investigation had an executive level position responsible for collecting and analyzing all of the dispersed and diverse data that were available and potentially relevant to the search? What if a multinational chief data officer (CDO) had been in place to manage all of the information that was available?

Companies have recognized the value of just such a position for some time. The first reported chief data officer was established in 2003 by Capital One Financial Corp., Yahoo, and Microsoft Germany were early adopters. In little over a decade, hundreds of organizations, including US federal and state agencies, have created chief data officer positions, although the jobs often are given different titles. In time, the initials CDO may become as familiar as CEO, CFO, and CIO.

Driving the trend is the phenomenon of big data—the explosion of information made possible by the great advances that we have seen in recent years in communications, computers, and storage.

Read the full post at Quartz.

Stuart Madnick is co-head of the MIT Total Data Quality Management and MIT Information Quality programs. He is also a professor at MIT Sloan School of Management.

How bad data can lead to good decisions (sometimes) — Roger M. Stein

MIT Sloan Senior Lecturer Roger Stein

MIT Sloan Senior Lecturer Roger Stein

From Computerworld

Before companies can profit from big data, they often must deal with bad data. There may indeed be gold in the mountains of information that firms collect today, but there also are stores of contaminated or “noisy” data. In large organizations, especially financial institutions, data often suffer from mislabeling, omissions, and other inaccuracies. In firms that have undergone mergers or acquisitions, the problem is usually worse.

Contaminated data is a fact of life in statistics and econometrics. It is tempting to ignore or throw out bad data, or to assume that it can be “fixed” (or even identified) somehow. In general, this is not the case.

Read More »