talent analytics: a systems perspective
TRANSCRIPT
Use of Analytics in Talent Decisions: Systems PerspectiveSharad Verma Indian Human Capital Summit Taj Land’s End, Mumbai 23 Jun 2016
Analytics in Talent DecisionsPoints to be covered today:
A systems Approach to Talent Analytics
Analytics as an Engine to Innovation
Real example of Human Capital Analytics (results from predictive retention analytics)
“In God we trust, everything else is data”Frequently heard at Google
Systems approach to Talent Analytics
What is a system and how does it work?
Goals versus systems
Components of a Talent Analytics system
Deterrents: why analytics fail to deliver results
What is a system A system is a way of doing things that gives predictable and improved results repeatedly over a period of time. It includes processes, people, tools, technology, mindset, habits, culture and environment.
Analytics can be used to build robust and high performance systems over time
Goals versus systemsGoal is a specific, measurable target to be achieved within a defined timeframe
System ensures delivering results repeatedly over time
EXAMPLES
Goals versus systemsGoal: Reduce employee attrition by 5% in the next financial year
System: Build a way to continuously assess potential attrition, identify drivers of both retention and attrition, inform decision making and take a series of corrective actions, measure effectiveness of actions, course correct (Talent Retention System)
Note: (An intelligent system self-corrects via feedback)
Goals versus systemsGoal: Achieve internal promotion target of 60%
System: Build a way to identify factors that prepare people for next level, predictably build those factors in selection, talent assessment and promotions, measure whether internal promotees are successful or not and reasons why (Talent Management System)
Note: (System may sound like “how” or process, but its more, it is a specific way of doing things to deliver results)
Goals versus systemsPersonal Goal: Lose 5 kg weight in next 2 months
System: Implement a way of living that reduces risk of disease and improves fitness
Note: (System may sound generic compared to a specific Goal but it is not - its success can be measured based on the results delivered - good or bad)
Components of a Talent Analytics System
People Technology
Culture Leadership
Processes
A BRESULTS
SMEs Context
Talent Analytics SystemsFor good outcomes, start with good questions
What qualities make excellent internal leaders? How do we spot and develop those qualities?
What are the predictors at the time of recruitment that someone will be a star performer? What are the right questions to assess those predictors?
What factors can accurately predict risk of attrition?
What factors result in consistent high performance on a specific job?
Talent Analytics SystemsAfter defining a clear desired outcome:
List hypotheses
Prepare datasets
Map current biases arising from experience, judgment or personal worldview
Use statistical modelling
Understand context
Draw from subject matter experts
Why analytics fails to deliver These org factors are deterrents to a data-driven culture:
Not knowing what to measure and “why” (ill-defined results and outcomes, bad start questions, presumptive approach)
Hierarchy
Experience
Bias (conscious and unconscious)
Leadership (can be an enabler also)
Judgment / beliefs / emotions
Skills
Tools
Lack of scientific and statistical temperament: desire for quick conclusions / reading data wrong
Analytics as an engine to innovation
> There are less than 50 companies that are true innovators and not more than 500 individual innovation / analytics influencers. In any field, less than 5% of companies/people fit this category
> Imitators repeat the work, ideas, practices of innovators many many times over
> Analytics need to be built keeping the context in mind - just repeating an isolated practice from Google may not give same results
Influencers Imitators
BUILD YOUR OWN ANALYTICS MODEL
Real life exampleA predictive model for talent retention (Financial Technology)
Over 2 years, a rigorously tested predictive analytics tool helped to:
- reduced attrition by over 12% annually
- significantly improved “managing for retention” ability
- delivered important lessons on drivers of retention and engagement - used by leadership, managers and HR
Some interesting findingsStarting question:
What is more important for retention managers or compensation?
Notes:
- Managers believe and repeatedly tell management it is compensation
- Leadership believes “people leave managers not companies”
Some interesting findingsStatistically valid findings:
- People who are unhappy with compensation are willing to wait for 6 months or more before they decide to leave
- People who are unhappy with managers leave within 3 months
Some interesting findingsStarting question:
What are the most important predictors of retention?
Some interesting findingsStatistically valid findings:
4 highest scorers and most consistent predictors of employee retention when together:
1. “I am in the right place” (environment, company)
2. “I am in the right job” (meaningful, challenging role)
3. “I have great relationships at work” (colleagues)
4. “I am treated fairly” (rewards, performance measurement)
Some interesting findingsOther important findings:
1. Work/life balance when positive is an “equalizer” - it compensates for low scores on role, compensation etc
2. Just scoring low in terms of satisfaction does not mean the employee will leave. - Unhappy employees will stay because of “equalizers” - in fact many tenured employees have lower happiness scores
3. A dissatisfied manager is more likely to have dissatisfied team and higher attrition rate
4. Work/life balance when negative is an immediate deal breaker
5. There are four SWEET SPOTS: a) a convenient commute time b) a great manager c) a perfect role d) great people to work with
6. Three most important qualities of managers - TRUST, HOPE and WORTH
Thank you Email: [email protected] Twitter: @iSharad Blog: iSharad.com