The amount of data that organizations have in their storage systems has increased in recent years from terabytes to petabytes and now to exabytes—with each exabyte equaling one billion gigabytes. The proliferation of data continues unabated. So, it’s with good reason that most c-level professionals are thinking about how to mine their data effectively. 

Data mining provides tremendous potential for gaining insight into customer buying patterns, internal efficiency improvements, employee engagement, and employee development opportunities, to name a few. While in the past data analysis supported strategic business planning on a relatively small scale, it is now leveraged by people at every level in every department. Learning and development professionals are no exception. 

Acceleration of Data 

Data has been on an incredible growth trajectory. Yet with the Internet of Things soon to move into broader adoption, the amount of data will accelerate even further. Sensor data will start to come in from people, systems, and smart objects of every kind. 

Data’s usefulness is also increasing. Data’s earliest applications were primarily collection and reporting. Then data became an enabler of analysis and decision support. Now, data is used to make predictions, based on everything from weather patterns to consumer buying preferences to healthcare. 

Learning and development organizations can benefit from advanced uses of data. Data analytics is helping to dynamically enable smart personalization of content to individual learner needs and channel preferences. Analytics are helping identify learning content that could become sources of revenue. Enterprise data is being integrated into learning applications to provide context rich learning environments. In countless ways, data has the capacity to add intelligence to the traditionally subjective nature of talent development. 

From Reports to Predictions 

Data is revealing new insights for organizations that choose to leverage it.  If you still fail to understand what all the fuss is about, make data mining, analysis, and analytics a priority in your organization. Move your data strategy from reporting to decision support and prediction. As you do, here are a few ideas to consider: 

Start your efforts by asking challenging questions that cover L&D needs, as well as concerns for specific lines of business. Some examples: 

• What skills have the most impact on new product innovation?

• What training offerings might we provide external customers?

• What training assets provide us with the most competitive differentiation?

• What sales training methods are most effective in reducing the sales cycle?  

• What development resources should be outsourced rather than provided in house?

Consider the potential of new databases. Expand your thinking from the data available in existing databases to new databases. Include external sources of data as well as combining multiple sources of internal data, such as employee performance data,  employee productivity data, career progression data, competency proficiency data, salary data and production metrics, first-year employee performance data, initial interview evaluations, job opening information, and industry skills gaps. 

Build cross functional teams of information technology and domain experts to analyze the data. Pair database experts with domain experts. These groups can work together to understand and connect data sources to decisions. 

Don’t assume all data is good data. Not all data is reliable. Don’t expect a precise answer when the underlying data isn’t precise. You may need to settle for ranges rather than specific numbers. Or, you may need to develop new sources of data. 

Challenge the results. Data correlation doesn’t prove causation. Just because umbrellas are prevalent when it rains doesn’t mean that umbrellas cause the rain. Do further research and analysis to confirm correlations and causations. Don’t blindly accept data based conclusions.  

Don’t dismiss instinct. Instinct has inherent biases, but it also provides common sense that data may not. Conclusions drawn from data need to pass the common sense test. 

Bottom Line 

Even when you have great data and accurate correlations, bad decisions still happen. Decisions are still postponed. Indeed, data analytics doesn’t guarantee great decisions. Remain vigilant about the importance of gaining buy-in for decisions, evaluating multiple alternatives, performing cost justifications, and creating detailed implementation plans. Crossing the knowing-doing gap will still be your organization’s biggest challenge.