Ten insights from IT leaders on the front lines of big-data deployments.
The IBM Center for the Business of Government recently interviewed 28 public-sector CIOs at the federal, state and local levels about the challenges of implementing big-data projects. The findings below are excerpted from a larger report, “Realizing the Promise of Big Data,” written by Arizona State University’s Kevin Desouza.
1. Public agencies are in the early days of their big-data efforts.
CIOs overwhelmingly report that they are just getting started with big-data efforts. While they see the value in synthesizing databases and using analytics, the challenge of big data has overwhelmed them.
Many CIOs also lack a tangible framework to guide their big-data efforts. The financial, policy, and pragmatic intricacies required to set up a big-data project are sometimes beyond the time constraints and expertise of many CIOs and their staff. In addition, the distinction between big data and conventional data is often not clearly understood.
All CIOs are aware of the value proposition of data analytics for their agency. While they do not doubt that big data could revolutionize service delivery and streamline business operations, CIOs believe they have “limited bandwidth” to invest in these efforts. CIOs are also aware that at the present time, much of the existing data in their systems goes unanalyzed.
2. Many CIOs fight the perception that big data is a passing fad.
CIOs say they have had to fight the perception that big data is just a buzzword or a fad. All CIOs note they have to tread carefully when discussing big-data efforts, and a large proportion even note that they have avoided using the term “big data” because of its negative connotation in their agency.
CIOs who had launched big-data efforts received support by focusing on the organizational and business value that would be gained from investments in data analytics. Technologies that would be used for data analysis were placed in the background and were never the focus of the conversation.
I do not have a single staff member who knows how to run simple regressions.... We can be the installers of tools rather easily, but then who is going to train employees [outside IT]?
3. Most CIOs are now primarily dealing with the issue of managing large volumes of data, integrating data across database systems and building an analytical capacity to mine data.
All of the CIOs interviewed are now dealing with structured data. None of the CIOs interviewed have begun to explore the intricacies of managing unstructured data.
Even when it comes to structured data, most CIOs admit that they are not dealing with datasets that are truly big. They are experimenting with creating executive dashboards and using information visualization techniques to make the large volumes of information easily accessible and comprehensible.
4. CIOs report that some big-data projects are now focused on streamlining business processes.
Some of the CIOs interviewed report that they are now focusing on identifying opportunities to streamline and re-engineer their business processes through investments in data management. These CIOs are focusing on process management as a vehicle to gain support from stakeholders, both internal and external.
5. CIOs do not anticipate significant investments in new technology for big data.
Eighty percent of CIOs interviewed report that they would not need to make any significant investments into technology, both hardware and software, during the first few years of their big-data programs. Instead, they would need to find more economical and strategic ways of deploying their current information technology assets.
6. CIOs report a need to bolster their human capital, including their analytical capability.
CIOs report that they are in the midst of bolstering their human capital through hiring new staff and training existing staff. In almost all cases, however, stagnant or diminishing budgets have severely impacted the capabilities of most public-sector IT units. CIOs report losing talented individuals to the private sector or early retirements. In addition, the inability to send staff for training and skill development has resulted in playing catch-up when it comes to big-data skills.
We should be in the business of data governance and data policing.... Our business units should be the ones in charge of analytics.
7. CIOs are now exploring approaches to data governance.
The issue of data governance has been the primary concern for CIOs. All CIOs note that most of the data residing in their information systems is not readily suitable for analysis. Most of the data lacks integrity and could not be easily integrated across systems due to a lack of standardization in data definitions, and even if data could be integrated, there are security and privacy considerations that need to be worked through.
CIOs say that poor data governance is the most critical factor holding up agencies in their efforts to pursue big data.
8. CIOs do not recommend IT units as owners of big-data projects.
CIOs caution against IT units being the owners or instigators of big-data projects. Instead, they believe that senior management support is necessary for success. While senior management support may be a requirement for most IT projects, CIOs say the complexity, transformational nature and upfront investments mean active senior management involvement is absolutely essential for big-data projects.
9. CIOs believe that collaborative leadership is crucial for the success of big-data projects.
CIOs report creating interdepartmental or interagency working groups for big-data projects. These working groups brought together key organizational stakeholders to move the project ahead.
CIOs also rely heavily on their professional networks as they traverse the uncharted waters of big data. Most CIOs report checking in with their peers in other agencies to seek out information and insight. Based on our limited sample of interviews, we did find that CIOs who are more connected to their professional communities (e.g., they speak at industry conferences, are named by other interviewees as standout exemplars, etc.) are further along in their big-data efforts.
We continue to build on a fragile foundation.... Data governance is not sexy and no one wants to do it, yet it is our Achilles’ heel.
10. CIOs are becoming champions of analytics and evidence-driven decision-making.
CIOs note that they are becoming the de facto champions for analytics and evidence-driven decision-making within their organizations. This is a role that many of the CIOs did not envision for themselves but one with which they are getting increasingly comfortable. Given that most public agencies have been influenced by the trends of open government, CIOs have had to become stewards and disseminators of data.
NEXT STORY: Defense Connect Online hits milestone