Why Analytics was born

Learning Analytics: Virtual roundtable with theory & practice

This virtual event took place on Friday, September 27, 2019, 1: 00-3: 30 p.m. - initiated by the “Expert Forum Digital Talent Management” initiative, which was brought into being by the Mannheim Business School, the University of Mannheim, SAP and tts has been.


We tried to organize interesting impulses from experts from business and science. It was about the possibilities to collect and evaluate data from learners, as well as to use the potential of this information in the best possible way. Here is a short summary including links. I put the recording here on Youtube.

Intro by Klaus Kräft and me with general theses on learning analytics. In the input (slides here) I concluded with a poll, as well as general theses:

  • The approach to learning must evolve from reporting to analytics
  • Objectives should be decision support and evidence rather than justification
  • A tip: focus on output & process KPIs instead of just input such as time or number of participants
  • The return on expectations along the stakeholder interest is at least as important as the ROI
  • Analytics help to measure impact via KPIs in an agile world, not years of studies


Digital traces of learning in the process of work

was the impulse from Prof. Dr. Andreas Rausch, Chair of Business Education, Learning in the Work Process at the University of Mannheim. The topic here was “Workplace Learning Analytics”. For me a forward-looking approach, albeit more difficult than analytics of formal learning. Concrete approaches such as the role of analytics in performance support or in learning in social networks were then discussed.

You can find the slides here.


Success conditions for the introduction of learning analytics

Professor Dr. Dirk Ifenthaler, Chair of Business Education - Technology-Based Instruction Design at the University of Mannheim outlined the topic and showed general framework conditions such as the result of meta-analyzes. In particular, he also pointed out that learning analytics can be used for better learner centering in order to better adapt learning to the individual. On slide 23, he also shows a nice grid with which you can assess yourself along the maturity level dimensions such as summative / real-time formative / predictive-prescriptive analytics. The slides can be found here.


Learning Analytics at SAP

Yikai Cao, Data Scientists for People Analytics Talent, Leadership, and Learning provided Learning Analytics at SAP from. In addition to the general approach, he showed two examples and experiences. Using a SAP Leadership Development Program, he showed how to proceed. Business & HR key figures (such as fluctuation) showed better values ​​compared to a group without the training. He also showed a tool that Learning Analytics uses to support HR professionals and managers at SAP. In the skills comparison you can see the development as well as the benchmark of different skills - be it with regard to importance, location or similar values. An exciting approach to specifically support the business. His slides can be found here.


Learning Analytics Analysis of educational activities at SBB

This lecture was about the reporting needs at SBB for the education department. Specifically, a demo was shown - the cockpit significantly optimizes quality assurance. You can find the slides here.



Learning Analytics is a broad field with little implementation, especially when it goes beyond simple reporting. What brings added value to companies can be quite different - this is shown by the various examples from SAP & SBB. Science can absolutely provide a framework, be it with evidence or meta-analyzes. In this particular case, science itself can be the subject of research when teaching and learning at universities is analyzed. Personally, I think everything is good that helps more learner centeredness - and there is a lot there.

The feedback on the session was good. Maybe it was a bit long at 2.5 hours and, as so often, we wanted to get a lot of content across. So we will certainly repeat a similar format, but then more with 2 instead of 4 lectures - and with more interaction. Zoom offers approaches such as breakout groups - we will certainly use even more simple approaches such as polls or chat. My personal advantage of this virtual event was that I could do it comfortably from home without being stuck in traffic for a long time.


We hope to have given one or the other inspiration. We'll be doing another event next year - at least a virtual one, maybe a physical one after all.

Comments or feedback are also welcome here in the comments.