Share on LinkedInTweet about this on TwitterShare on Google+Share on FacebookShare on Reddit

 Talent Analytics bigTalent Analytics tag

Organizations often attribute success directly to “their most important asset,” their people. So, it is a welcome trend in analytics to see more and more people, events and publications talking about talent and workforce analytics while more companies have begun to look inward at their own people as a data source for performance-related analytics.

Indeed, just in the last week, you may have read about talent analytics-related topics in Information Week, the The Atlantic, CIO Review Magazine, or a myriad of other resources posted daily on our twitter feed.

Perhaps you even read about using talent analytics to reduce call center turnoverat a major financial company or Accenture’s work to enhance recruiting and retention.

While talent or workforce analytics seems to be everywhere you look, analysts and data scientists working on employee performance optimization efforts today have some fundamental questions, including:

  • Hype aside, what is involved in the actual exercise of this kind of analytics?
  • How do you measure people?
  • What do “raw talent” metrics look like, and how are they processed?
  • How are advanced analytical methods used with these numbers to categorize groups and predict outcomes?
  • Do regressions, clusters, trees, other methods play nicely with people data? Most importantly, do they work?

On Tuesday 11/26 from 12-1pm Eastern, Talent Analytics Chief Scientist Pasha Roberts will answer these questions and provide technical details for working with the unique and high value “raw talent” dataset. This is an ideal opportunity to cut through the hype and get a pragmatic, technical perspective from a true industry-leader on using predictive talent analytics to optimize talent.

Who Should Attend?

Analysts, Data Scientists, Analytics Consultants, anyone with Workforce/Talent Analytics in their title, Risk Consultants, CTOs, CFOs – anyone charged with finding ways to use employee data to predict outcomes.

Register now.

11 Reasons to Attend ‘Predictive Talent Analytics: Technical Details’ on 11/26

  • Learn who, when, and why to measure “raw talent” information – for employees, job seekers, customers, and more.
  • Evaluate several sampling techniques, useful in different client situations.
  • Understand the specifics of the “raw talent” dataset, without the psychological jargon.
  • Expand your sense of “performance data” to include a wide range of outcomes.
  • Focus stakeholders to define quality performance metrics as independent variables.
  • Learn strategies to cope with non-existent or ambiguous performance metrics.
  • Understand the unique challenge of measuring turnover or attrition.
  • Learn how to apply clustering techniques to find unseen groupings within performance, descriptive or talent data.
  • See the results of regression, decision tree, random forest, and other techniques, predicting performance from raw talent data.
  • Evaluate the lift from these predictive models to estimate the benefits of model deployment.
  • See how to deploy a finished model into the Award-winning Advisor platform.

If you want to learn about predictive talent analytics, you won’t want to miss it!  Register now.

Mike Kennedy is an Analytics Evangelist at Talent Analytics, Corp. He can be reached via mike[at]talentanalytics.com and @talentanalytics.

 

Share on LinkedInTweet about this on TwitterShare on Google+Share on FacebookShare on Reddit

Blog Publisher / Head of Data Science Search

Founder & Head of Data Science Search at Starbridge Partners, LLC.