What’s the Big Deal About Big Data?
Big data can be daunting, but its analysis is making significant differences in organizations. By analyzing complex data sets across functional silos, organizations are gaining insights to help catalyze change, improve access to experts, speed onboarding, retain talent, and identify root causes for complicated issues. It improves the learning environment, and even the Learning & Development organization itself.
The people who drive value for an organization aren’t necessarily those in authority on the formal organizational charts. They often are those with depth and breadth of expertise, who influence others, know how the organization really works, and can reach beyond silos to accomplish results.
Water engineering firm MWH Global, for example, used big data analytics to identify the company’s top collaborators and then deployed them as catalysts to help consolidate activities as the company transitioned from a function-based IT structure to a shared services model. “The company identified the top change agents and publicly recognized them as role models. After six months, MWH Global saved $25 million,” recounts Cecyl Hobbs, SVP, Business Development and Marketing, at social network analytics company Activate Networks. By improving access to internal experts, the company was able to overcome bottlenecks and barriers more quickly than otherwise would have been possible and distribute information more effectively throughout the network.
Halliburton worked with Activate Networks to improve communication among its global sites when a network analysis showed multiple clusters with few ties among them. Based on that analysis, Halliburton began strengthening cross-platform ties by creating mixed project teams, rotating well-connected individuals to other platforms, and creating an electronic expertise locator. Nine months later, connections had increased 25 percent and operational productivity 10 percent, costs caused by poor quality were slashed 66 percent, and customer dissatisfaction decreased 24 percent. New product revenue increased 22 percent. The improvements were attributed to the ability to make shared decisions more efficiently and to exchange best practices and innovations.
Other organizations use big data analysis to retain talent. “Engagement, performance, and social connectivity are key elements of flight risks,” Hobbs points out. “Are employees sought out for their expertise, considered critical to a project, sidelined, or overloaded? You can look at this over time and understand where an individual fits. Those who are becoming more isolated and less energized may be flight risks.”
For onboarding, the extent of individuals’ networks is the key to their success. “If a consultant isn’t well-integrated within 30 days, it’s a cause for concern.” Hobbs gives corporate execs a bit longer. Within 60 days, he says, it should be evident whether new executives are working closely with the necessary people and departments. “If that’s not happening or if the network is lopsided, that’s a key indicator the executive isn’t thoroughly onboarded.”
Defense Acquisition University (DAU) takes a different approach, using big data analytics to gauge the effectiveness of its learning programs. DAU provides training for more than 151,000 active and reserve procurement and IT personnel throughout the U.S. military. Recognized as the best corporate university of 2013 by the Global Council of Corporate Universities, DAU worked with Knowledge Advisors to integrate data from multiple systems—including human resources, budgeting, and accounting—with learning databases and student information. Consequently, “we can benchmark against a decade of surveys, looking at courseware and facilities, and how individual courses affect the organization’s performance,” says Dr. Chris Hardy, director of Strategic Planning and Learning Analytics, DAU.
Hundreds of thousands of post-training surveys on course quality and instructor effectiveness are completed each year immediately after course completion and 60 days later to assess the effect of the course upon students’ job performance and business outcomes. Key findings indicate that courseware quality is more important for younger learners—who prefer e-learning—and for those with some graduate-level education, than for older learners, who prefer traditional classrooms and effective instructors. Instructor effectiveness was earmarked for improvement because analysis revealed “a huge relationship between instructor effectiveness and courseware quality.” By comparing benchmarked data, Dr. Hardy learned that DAU instructors are more influential at DAU than at other organizations. Guest speakers triggered higher levels of individual learning, which were reflected later in job impact and business results.
Dr. Hardy is advancing DAU’s capabilities with a new learning analytics team that performs trends analysis across business lines. “Looking reactively, you don’t see the trends,” he says. “But when analyzing data for things such as graduates vs. return rates, perceived course quality, course location, business unit differences, etc., trends become evident. Then, their root causes can be identified and any issues can be addressed. For example, we used Knowledge Advisors’ Metrics that Matter software to analyze why distance learning return rates were dropping. We learned the government shutdown and furloughs had lowered morale,” so students weren’t completing surveys or attending classes.
Currently, Dr. Hardy says, “we’re connecting the system to the student information system to link business outcomes to training.” DAU already tracks training locations, quality, costs, student evaluations, and applicability to the job. When finished, the linked system will operate like a talent management system for learning, linking to knowledge-sharing systems with features such as templates, regulations, and just-in-time training.
LEVERAGING HIDDEN INFORMATION
Sophisticated analytics capabilities are the key to unlocking the information buried in data that organizations already have but either aren’t using or don’t realize they have. This approach to big data analysis combines network science and behavioral science to improve collaboration and employee engagement. As Hobbs elaborates, “we gather information to identify networks, individual influence in the community, and the effects on the group. We’re using scalable solutions to give both a micro and macro view of key professional relations.” Influence isn’t necessarily a function of authority, he points out.
Activate Networks’ Activate Social Platform for Enterprise software solution can map networks from millions of individuals. For example, it aggregates and analyzes the metadata and header information from e-mail traffic, including the sender/receiver and time stamps (but not the content of the e-mail) to identify individuals’ communication networks. “By running advanced analytics, organizations can get qualitative insights that identify the information brokers and the information bottlenecks,” Hobbs explains.
When identifying the information brokers, the company “builds a profile of descriptive data, such as location, gender, and tenure, and then layers on behavioral information, including their network and e-mail data, engagement, and additional skills. The result pinpoints the real energizers who empower people in an organization,” Hobbs says.
Understanding those relationships “drives time to market, simplifies organizational complexity, enhances collaboration, minimizes predictable errors, and helps organizations monitor the results of changes over time. That, in turn, can accelerate revenue growth by shortening sales cycles, and generating warmer leads and a seamless customer experience for internal and external customers,” Hobbs says.
But even with advanced analytics, some data can remain unreachable. Physician narratives in medical records are a good example. These narratives are critical, particularly in difficult or chronic cases, yet require natural language analytics to unlock the information so it can be applied to other subsets of patients. Donald Farmer, VP, product manager, Qlik Technologies, calls this “water cooler collaboration,” because it presents information in a way that mimics how the human mind naturally absorbs data.
“A question is never just technical. Humans share through dialogs and stories,” Farmer says. “Natural analytics is a combination of technologies and experiences that builds upon cognitive techniques and innate skills,” and, therefore, can leverage value from those stories.
Ultimately, big data analytics will increase in value as organizations deploy them to make cross-functional connections. These will foster insights that address the heart of issues that affect learning and, thus, productivity and profits. And that is a big deal.
DO COMPANIES HAVE BIG DATA SKILLS?
Only 1 in 4 organizations indicated they have an ability to meet their analytics needs, while another 17 percent plan additional hiring to do so, according to the American Management Association’s global survey of 800 respondents from more than 50 industries conducted by the Institute for Corporate Productivity (i4cp). The majority of respondents (47 percent) plan to invest in training to meet their capabilities gaps. Human Resources and Sales are seen as lagging in analytical skills when compared with other organizational functions. The survey found that lack of resources and corporate culture are the biggest impediments to an organization’s ability to leverage big data.
HR professionals have a critical role to play in creating and shaping the new analytical workforce, notes i4cp’s report, “The Age of Big Data: A Progress Report for Organizations and HR.” Here are some lessons learned from today’s market leaders on how to get started, as reported by i4cp’s Cliff Stevenson:
- IDENTIFY ANALYTICAL NEEDS IN YOUR ORGANIZATION. Assess your workforce for analytical capabilities and use that data to determine where to focus first. Any departments that fall well below where the acceptable level is should be dealt with first, but if all else is equal, work on increasing the analytical abilities of top leaders either through executive development or recruitment.
- BUILD ANALYTICAL STRENGTH. To build analytical acumen, training should focus on using data to make better decisions rather than on specific tools and data-crunching techniques—although those are still important for some jobs. This type of training will help employees approach problems from a more empirical point of view. Some functions within your organization already may have the needed skills and can be tapped as subject matter experts to help educate others.
- PREPARE TO MANAGE THE FLOW OF BIG DATA. The hubbub regarding big data is mostly about that first word: big. If organizations are planning on making use of the enormous data sets available to them, infrastructure must be in place beforehand. Enterprise-wide HRIS may or may not be able to leverage the massive amounts of data collected, so it’s important to understand what you are hoping to find before plunging into the overwhelming current of big data.
- EMBRACE THE ANALYTICAL DECISION-MAKING MINDSET. Changing from an instinctual, experience-based decision-making organization to a data-driven one isn’t as simple as increasing your organization’s analytical abilities. The very way in which problems are viewed has to be changed, which is why it is so important to have leaders who understand and use data-based/evidence-based decision-making. Merely having more data accomplishes nothing if that data isn’t used to make better, more fact-based decisions.