This year’s annual MIT Chief Data Officer and Information Quality (CDOIQ) Symposium was held on July 17 to 18, 2013 at the MIT campus in Cambridge, Massachusetts. This event brings together Big Data practitioners to discuss the real need to improve overall data quality in the quest to leverage big data demands.
theCUBE was present for the event and interviewed a number of people over the two day period. For CIOs, clear trends emerged, some of which could be good and others not-so-good for the CIO profession. In this professional alert, I will discuss three significant data concerns that need to be addressed in organizations wishing to move into the world of Big Data and information analytics.
The volume of data continues to increase
During the CDOIQ Symposium, theCube interviewed Michael Nix, Director of Analytics at Fletcher Allen Health Care. Nix talked about the problem of managing the huge increases in data volume Big Data is driving. He uses an example from his own field: health care. Today, electronic medical records are big business, and the results are Big Data. Health care organizations are creating vast quantities of data that needs to be appropriately managed.
But healthcare isn’t alone; it’s but one industry experiencing a surge in data volume. Today, growing quantities of data are appearing just about everywhere, thanks in part to what has been termed the “Internet of Things”.
The rise in volume of data shouldn’t be seen as a problem, but as an opportunity. In what may first appear to be a haystack of data could lie a transformational needle awaiting discovery. Today, the job is to find ways to leverage these data assets, and that takes the right tools, the right talent, and the time to do so.
Data quality is an industry-wide problem
Perhaps the biggest challenge in these efforts is data quality. For a long time, I’ve lamented the poor quality of data present in many organizations. Organizations continue to struggle with ensuring that they get both good and clean data. “Good data” means that whatever information is being captured has relevance to the business. “Clean data” means that information is free of errors. Getting clean data is a matter of ensuring that organizations have good data entry processes and procedures and might entail an initial scrubbing to get data to a consistent state after which it can be considered “clean” again. For organizations that undertake a data cleansing effort, it’s critical to identify the original root cause of the dirty data and implement corrective policies and procedures to prevent cleaned data from drifting back to the muddy side of the data river.
Good data, on the other hand, can be harder to achieve. An organization needs good data to make sure that business decisions are made with insight. In fact, good data is critical to our very way of life. Justin Magruder, President and Senior Managing Director at Noetic Partners, says, “Data quality, and more specifically the lack of data, was a fundamental component of the financial crisis.” With better data, financial institutions can better manage risk, and individuals can better understand their own exposure in certain transactions. Data quality is at the heart of whether or not data is useful to an organization. In fact, I’m of the opinion that bad data can be worse than no data sometimes. At least with no data, experienced pros can make “gut” decisions rather than making bad decisions based on faulty data.
Getting to “good data” requires organizations to ensure that they are collecting the right information and doing so in a consistent way. This means that regular attention must be paid to ongoing high-level data needs across the organization. Failure to achieve data quality goals leads to decisions being made on insufficient information.
“Good data” can be difficult to achieve. It requires someone who understands what data is important – or could be important – to the organization and makes sure that data is collected in an actionable form. The next section discusses the person or division who might be responsible for these activities.
Does the rise of the CDO spell the fall of the CIO?
For the CIO profession itself, the Symposium discussed the rise of the Chief Data Officer (CDO), with arguments made for the addition of a CDO in any organization that relies on quantities of data in operations or for revenue purposes. Mario Faria, Big Data Technology Advisor at Technology Partner Network of the Bill & Melinda Gates Foundation, went so far as to say, “When you bring in a CFO, do you make a business case to bring your CFO?” Faria believes that the importance of the role should be clear based on the need to create value from the treasure troves of data many companies already possess. Faria also believes that the CDO should not report to the CIO as “Data should be a business issue, not a technical issue.” While I do not believe that a CDO position should simply be an automatic addition to the payroll, I do believe that data analytics is incredibly important to an organization.
Personally, I continue to believe that, if the CIO is seen primarily as a technical resource and not as a business resource, something has gone wrong in the organization or the organization has bestowed the CIO title to someone who is really a technology director. Unfortunately, for a wide variety of reasons, CIOs continue to struggle in many ways to be seen as equal to other C-level counterparts, and many continue to be regarded as glorified technicians rather than business equals who play a part in overall business strategy.
All that said, organizations must find ways to leverage their data to better effect. This will require the addition of resources that may or may not fall under the CIO’s purview. For organizations considering the addition of a CDO, a cost/benefit analysis should be performed and the executive team, including the CEO, should consider the organizational location for such a position, whether that means that the CDO reports to the CEO, to the CFO, COO, or CIO.
Action Item: Organizations must find ways to leverage their data to better effect. This will require the addition of resources that may or may not fall under the CIO’s purview. For organizations that are considering the addition of a CDO, a cost/benefit analysis should be performed and the executive team, including the CEO, should consider the organizational location for the position, whether that means that the CDO reports to the CEO, to the CFO, COO, or CIO.
For years, I’ve advocated that CIOs jettison as much baggage as possible in order to better focus on the business. The industry is approaching a point at which information needs will ultimately be addressed – with or without the CIO's involvement.