ASTD TK13: Ellen Wagner on Analytics

My notes from Ellen Wagner’s presentation introducing big data to ASTD.

• We’ve accumulated 2.8ZBs of data. That’s an insane amount of information.

• People are willing to give up some privacy for convenience – tracking on sites like Amazon, Netflix – all recommendations based on “data breadcrumbs” you’ve left behind.

• Learning organizations need to be more like businesses – looking at data, metrics, optimization.

• Enterprises must do what’s in the best interest of their stakeholders/shareholders.

• Analytics gives us the information we need to show value at the c level, demonstrate accountability.

• We can, and should, learn from trends in business optimization.

• eLearning Optimization – measure what we’re doing. We have to show what we’re doing or we won’t survive.

1. Measure
2. Execute
3. Automate
4. Extend
5. Innovate

• We now have expectations of accountability, transparency and quality.

• Question: how do we speak in the language of the business?
  
  Answer: we now need to be authentic – okay to speak like L&D if we have metrics to back it up. Ask the right questions – what are we looking for ?

• Even if you think you have silos, the data probably shows differently and exposes the connections. “All analyses and stakeholders are interconnected.”

• Where do you begin? No best practice yet. Folks are worried about what the data will show/expose.

• Data driven decision-making is aimed at making better decisions.

• Data tells us what happened and improves strategic planning.

• Model building is an iterative process. 70-80% of efforts should be spent on data exploration and understanding.

• Use data to improve decision making:

1. Collect it
2. Predict
3. Implement
4. Monitor
5. Decide
6. Refine
7. Start over

• Where is your data?

  ERPs, LMSs, HRIS, Surveys/end of course data

• What are people actually using and how? Not just launches and completions but how people are engaging with the content. Figure that out and you’ll have a better idea of what people want, need and what you should spend money on.

• Lots of talk about evals, but we rarely actually do it or do it well.

• Get it at the front of the convo instead of at the end.

• LMS are messy – never built for the data we now want.

• Analytical solutions:
  Don’t buy the term, make sure you know what you want, and what you want to do with the information, and use features or request features that allow you to do that.

• Start simple – Google Analytics, use that data to show value in data collection and get more/better analytics.

• Look at marketing and advertising. How they use metrics and tracking. Learn from them.

• Adobe now using LearnBench – tracking data.

• Creation of middleware db should be priority for all institutions.

• SQL or Hadoop for unstructured

• SPSS or SAS for structured

• People resist new ideas if it clashes with what they already know. Be prepared for that.