A Little Social Learning and Predictive Analysis
I recently returned from the spring CLO Symposium. The theme was “Learning Evolution: Alignment, Agility, and Adaptability.” There were approximately 350 attendees and the workshops covered:
- Organizational hacking
- Social learning
- Adaptive design
- Predictive analytics
- Business acumen
- Developing business advisers
- Social learning
- Adapting training to worker skill realities
- Lessons from the LearningElite
- Predictive analytics, and
- Social learning.
Did I mention topics included predictive analytics and social learning?
Of course, I’m exaggerating but measurement, specifically anticipating need, and social learning were hot topics. Even if they weren’t included in every workshop, attendees were discussing them during breaks and in the evening. Both have great appeal. I think everyone would like to predict the future, especially when it comes to training demand and organizational impact. And, social media has everyone’s attention. I don’t have to tell you how many people are on Facebook, Twitter, or LinkedIn. We learn so much from each other and robust search capabilities. Of course, we’d like to bring pull-based, experiential learning into our organizations.
Predicative analytics is a phrase that deserves some definition before it’s discussed. At its simplest, a person might mean anticipating a training requirement based on some organizational change. At its most complex, and strictly defined, it includes methods from statistics, data mining, and game theory to analyze historical data and predict the future. The simple definition is executed when a learning leader recalls the chaos following a new product roll-out and implements training to ensure success. The strict definition has created a strategic advantage for Wal-Mart as they manage inventory levels and distribution.
Most organizations we work with lack the analytical sophistication and data necessary to conduct predictive analysis.
There are ways to develop analytical sophistication. You can:
- Purchase software and attend the training.
- Hire people with the know-how.
- Identify people with an aptitude for analysis and create a team.
…But, it is nearly impossible to close the data gap.
Predictive analysis is based on historical data. Few learning organizations can determine who attended which interventions much less how attendance correlates to the internal or external environment. You have to understand what happened to anticipate what will happen.
I recommend implementing a data collection process and working to optimize ongoing interventions as a first step. As your analytical sophistication grows, you’ll have the data to support predictive analysis.
This “walk before you run” approach shouldn’t be discouraging. Conversations at the symposium indicate many people are using the “predictive analytics” term but few are applying it beyond the simple definition above. In time, I’m confident we’ll be better able to anticipate the learning needs of an organization but it’s especially challenging because we’re dealing with people. People often act in unexpected ways. Many attempts at prediction are not very successful despite enormous volumes of data, unprecedented calculatory power, and incredible incentives to succeed. Consider the stock market.
Social learning might be a softer target. Numerous organizations are implementing social learning platforms. We have one here at Bellevue University. We’ve had great success with content and expertise location. As you might imagine, I was very interested in the organizational impact of the system. The Lab identified several key performance indicators related to social media and developed measurement tools. The methodology isn’t ready to share and we don’t have results yet but we will by mid-year. When it’s complete, we will publish a case study and share the information at the summer Colloquium in Omaha.
It was a great symposium. I learned a great deal and had the opportunity to visit with numerous thought leaders. I believe the future holds great things for learning leaders and their organizations. Loosely defined, I guess that’s a little social learning and predictive analysis.