Talent Analytics
The elixir to a successful HR career
Feb 2012 - NHRD
Do You….


• Often catch yourself saying ―I feel this is the right thing to do‖?

• Still use interview conversations to predict the suitability of a potential
  employee with a reasonable degree of certainty?

• Track attrition statistics on percentage points and feel happy if it shows
  a downward trend?

• Have a plethora of training programs but no clear link to business
  impact?

• Want to make an impact but did not know how to do it?

…….And, If I wanted to lift my sagging career, what would I do?



1   Footer                                             Copyright © 2011 Deloitte Development LLC. All rights reserved.
Moneyball…….


• What does one needs to do when one does not have the money
  muscle to get talent ?

• Has a boss who shows no support for any expense !

• The business and your self respect depends on an outcome that
  seems challenging to achieve?

• And you wish to make a difference !!

• THINK DIFFERENTLY and use WHAT is AVAILABLE to you…DATA




        In God we trust, all others bring Data – Edward Demming

2   Footer                                       Copyright © 2011 Deloitte Development LLC. All rights reserved.
Talent analytics – your barometer



                                          Workforce planning and deployment
             Talent management elements



                                             Talent sourcing and selection


                                            People/leadership development


                                             Performance management


                                             Rewards and recognition


                                              Knowledge management




3   Footer                                                    Copyright © 2011 Deloitte Development LLC. All rights reserved.
Weather reports – track the talent radar

                                      • Those which summarize and compare
                                       operational and/or financial data on key
                                       workforce variables within defined time
                                                       frames.
                                     – e.g., totals, averages, percentages, and trends.

                                                        Today




                                                                                              • Mathematical models that
                                                                                                use multiple internal and
                                                                                                 external data sources to
                                                                                               predict future talent events.
    • Those which apply one or                                                                     – e.g., a predictive model
     two internal data sources to                                                                      that uses internal and
                                                       Talent
      derive useful information.                         …                                            external employee level
                                                      analytics                                          data to predict the
       – e.g., the past education
                                                                                                    likelihood that a particular
              experience of job
                                                                                                    employee will resign in the
        candidates is compared to
                                                                                                        next six months and
       job performance during the
                                                                                                    supply the reasons for the
         first year of employment.
                                                                                                        prediction (e.g., long
                                                    Maximize
                                                                                                             commute).
                                                   performance
                        Yesterday                                                    Tomorrow



4       Footer                                                                       Copyright © 2011 Deloitte Development LLC. All rights reserved.
Back to
Management reports                                                                                           Talent
                                                                                                            Analytics




Attrition waterfall
                                                  2,161               198                             9,642

                10,000
                          8,480
                                          801


                8,000



                6,000
    Headcount




                4,000



                2,000



                         FY11     Attrition     New starts      New starts                            FY12
                         March                                   attrition                            Feb




5          Footer                                            Copyright © 2011 Deloitte Development LLC. All rights reserved.
Back to
Management reports (cont.)                                                                                                                      Talent
                                                                                                                                               Analytics




Performance rating movements

                                                                     1
                                                                   0 | 0%                 2
                                                                                       0 | 0%                        3
                                                                                                                  0 | 0%



                                     0 | 0%                                                         0 | 0%

                                              0 | 0%                                   0 | 0%

                                                                Outstanding


                                              Very                                     Very
                                              good     0 | 0%                 0 | 0%   good

                            Good     0 | 0%                                                      0 | 0%          Good
                   0 | 0%                                                                                                     0 | 0%
         Average                                                                                                                            Average

FY09       000              000               000                  000                  000                       000                           000
FY08       000              000               000                  000                  000                       000                           000

                            0 | 0%                                                                               0 | 0%

                                                                                                                                 Avg.     Avg.
                                                                                                                                    0 | 0%




6      Footer                                                                                   Copyright © 2011 Deloitte Development LLC. All rights reserved.
Weather reports – track the talent radar

                                      • Those which summarize and compare
                                       operational and/or financial data on key
                                       workforce variables within defined time
                                                       frames.
                                     – e.g., totals, averages, percentages, and trends.

                                                        Today




                                                                                              • Mathematical models that
                                                                                                use multiple internal and
                                                                                                 external data sources to
                                                                                               predict future talent events.
    • Those which apply one or                                                                     – e.g., a predictive model
     two internal data sources to                                                                      that uses internal and
                                                       Talent
      derive useful information.                         …                                            external employee level
                                                      analytics                                          data to predict the
       – e.g., the past education
                                                                                                    likelihood that a particular
              experience of job
                                                                                                    employee will resign in the
        candidates is compared to
                                                                                                        next six months and
       job performance during the
                                                                                                    supply the reasons for the
         first year of employment.
                                                                                                        prediction (e.g., long
                                                    Maximize
                                                                                                             commute).
                                                   performance
                        Yesterday                                                    Tomorrow



7       Footer                                                                       Copyright © 2011 Deloitte Development LLC. All rights reserved.
Back to
Employee commitment                                                                                                   Talent
                                                                                                                     Analytics




       Human Resource (HR)/                 Culture           Employee                                   Switching
      people practices (Drivers)          (Moderators)       commitment                                   factors
                                                           (Goal/Outcome)                              (Commitment
                                                                                                        moderators)
               • Service leadership
           • Change management
       • Employee communications         Employee Focus
          • Immediate supervision                              Employee
    • Employee growth and development                         Commitment                    Switching Costs
          • Training and education       Customer Focus        • Affective
          • Performance evaluation                             Attachment
         • Recognition of employee                           • Willing to be
                performance              Community Focus         Proactive
                                                                                      Switching Alternatives
                • Compensation                              • Intent to Stay
                   • Benefits
                 • Career-life fit       Financial Focus
      • Teamwork/team management
          • Diversity management
    • Customer relationship management




8     Footer                                                          Copyright © 2011 Deloitte Development LLC. All rights reserved.
Weather reports – track the talent radar

                                      • Those which summarize and compare
                                       operational and/or financial data on key
                                       workforce variables within defined time
                                                       frames.
                                     – e.g., totals, averages, percentages, and trends.

                                                        Today




                                                                                              • Mathematical models that
                                                                                                use multiple internal and
                                                                                                 external data sources to
                                                                                               predict future talent events.
    • Those which apply one or                                                                     – e.g., a predictive model
     two internal data sources to                                                                      that uses internal and
                                                       Talent
      derive useful information.                         …                                            external employee level
                                                      analytics                                          data to predict the
       – e.g., the past education
                                                                                                    likelihood that a particular
              experience of job
                                                                                                    employee will resign in the
        candidates is compared to
                                                                                                        next six months and
       job performance during the
                                                                                                    supply the reasons for the
         first year of employment.
                                                                                                        prediction (e.g., long
                                                    Maximize
                                                                                                             commute).
                                                   performance
                        Yesterday                                                    Tomorrow



9       Footer                                                                       Copyright © 2011 Deloitte Development LLC. All rights reserved.
Back to
Retention tracker                                                                                                                                         Talent
                                                                                                                                                         Analytics




Lift chart – demonstrates the effectiveness/benefit of the retention model

                               40

                               30

                               20                                                                                                              Extremely high
                                                                                                          High attrition risk
     Relative attrition risk




                                                                                                                                                attrition risk
                               10                                                                             segment                             segment

                                0

                               -10
                                               Low attrition risk       Moderate attrition risk
                               -20                segment                    segment

                               -30

                               -40
                                           1           2            3   4           5             6   7      8               9              10
                                                                                Deciles


                                                                            Model Equation

                                target = a+b1(tenure)+b2(commute to work)+ b3(pay)+ b4(rating)+b5(training hours)+…

                                                  This equation is used to give a score to each employee.



10                        Footer                                                                          Copyright © 2011 Deloitte Development LLC. All rights reserved.
Back to
Retention tracker (cont.)                                                                                                            Talent
                                                                                                                                    Analytics




Solution set to provide an overview of the attrition risk in the organization


                                     Organization
                                         4%
                                                                      Attrition risk projections can be analyzed
                                                                                 at organizational level

                       Business1     Business 2     Business 3
                          3%            2%             4%
                                                                      Attrition risk projections can be analyzed
                                                                                   at a business level

              Team 1           Team 2         Team 3         Team 4
               1%               2%                4%             5%
                                                                      Attrition risk projections can be analyzed
                                                                                     at a team level


                                                                      Attrition risk probabilities are generated
                                                                       for each employee in the organization




11   Footer                                                                          Copyright © 2011 Deloitte Development LLC. All rights reserved.
Weather reports – track the talent radar

                                    • Those which summarize and compare
                                     operational and/or financial data on key
                                     workforce variables within defined time
                                                     frames.
                                   – e.g., totals, averages, percentages, and trends.

                                                      Today




                                                                                            • Mathematical models that
                                                                                              use multiple internal and
                                                                                               external data sources to
                                                                                             predict future talent events.
 • Those which apply one or                                                                      – e.g., a predictive model
  two internal data sources to                                                                       that uses internal and
                                                     Talent
   derive useful information.                          …                                            external employee level
                                                    analytics                                          data to predict the
     – e.g., the past education
                                                                                                  likelihood that a particular
            experience of job
                                                                                                  employee will resign in the
      candidates is compared to
                                                                                                      next six months and
     job performance during the
                                                                                                  supply the reasons for the
       first year of employment.
                                                                                                      prediction (e.g., long
                                                  Maximize
                                                                                                           commute).
                                                 performance
                      Yesterday                                                    Tomorrow



12   Footer                                                                        Copyright © 2011 Deloitte Development LLC. All rights reserved.
Talent Analytics – A Model



• What is the business issue that we
  are looking to address?

                                       • What kind of data is required to
                                         analyze the issue?
                                       • How do we look at the data to draw
                                         meaningful conclusions?


• How are we measuring the
  effectiveness of these
  measures?
• Can we measure it in $ terms
  and report to the business?


                                       • What insights can we draw from the
                                         analysis?
                                       • Is it possible to extrapolate past
                                         data to predict future outcomes?




 13   Footer                            Copyright © 2011 Deloitte Development LLC. All rights reserved.
Analytics - The Past and The
Future
Evolution of Analytics



                                                                      Predictive
                                                                      Analytics


                                                        Cross
                                                     process and
                                                      functional
                                                       Analytics


                                          Basic
                                         Analytics



                          Consolidated
                           reporting


              Data and
                Basic
              Reporting




15   Footer                                                        Copyright © 2011 Deloitte Development LLC. All rights reserved.
Predictive Modeling Defined
Predictive modeling uses available internal and external data to predict future events
at an applicant, employee, or claimant level. Models can be designed to predict a
variety of outcomes and have broad based applications.



     Data Mining — a process, which utilizes a number of
     mathematical techniques, to analyze large quantities
     of internal and external data, in order to unlock
     previously unknown and meaningful business
     relationships.




     Predictive Modeling — the application of data mining
     techniques and algorithms to produce a
     mathematical model that can effectively predict and
     segment future events.




16    Footer                                                Copyright © 2011 Deloitte Development LLC. All rights reserved.
The 4 Critical Modeling Questions

1.     Is there a compelling problem or opportunity?
      •       What is the business case scope and size?
      •       What are the costs, including opportunity costs?


2.     Do we have the data we need?
      •       What is the state of available data and our ability to access it?


3.     Can we segment or predict potential outcomes and does it put
       us in a position to make a difference?
      •   Is there a basis to build a predictive model?
      •   Will the model output help us solve the defined business problem?


4.     Can we effectively act upon the predictive model output?
      •   What is our change readiness and ability to implement?
      •   Will implementation drive the required ROI?
      •   What is our political and legal climate?


17   Footer                                                            Copyright © 2011 Deloitte Development LLC. All rights reserved.
What Can Predictive Modeling Do?

     • Predictive models use available internal and external data to predict the
       likelihood of future events at the customer or employee level
     • Models are deployed by businesses to direct limited resources to the actions
       that will yield the largest economic benefit


For example…                                   Similarly…


     A business unit uses predictive            HR can use predictive models to
     models to maximize ‘customer               maximize ‘employee lifetime
     lifetime value’                            value’
        •       Targeting new customers           •   Recruiting and hiring new
                                                      employees
        •       Optimizing pricing, customer
                service and costs                 •   Optimizing development and
                                                      performance
        •       Focused customer retention
                                                  •   Focused employee retention




18     Footer                                                  Copyright © 2011 Deloitte Development LLC. All rights reserved.
Talent Acquisition – A Case in Point
 Problem Statement: How do we review more qualified candidates, faster, with
 improved accuracy, and with less cost?


                            150,000 applications p.a. which
                            results in an eligible candidate
                                     pool of 81000



                               Initial screening brings
                                this number down to
                                          8000



                                   Further filtration
                                   occurs through
                                   multiple rounds
                                    of interviews




                                    2100 Offers /
                                   1675 accepted

19   Footer                                                    Copyright © 2011 Deloitte Development LLC. All rights reserved.
Traditional Recruiting Data

• Who would be the most successful and Who would be the long term employee?


                    Sachin                            Saina                                          Rahul
          10 years of work experience       10 years of work experience           10 years of work experience
         4 previous employers in past 10   2 previous employers in past 10     1 previous employer in the past
                     years                             years                              10 years
            Current employer is large         Current company is small         Current company is a microchip
              technology company                technology company                      mfg. company
           Attended Tier I Engineering      Attended Tier II Engineering         Attended Tier III Engineering
                    College                          College                               College
               B.E. in Electronics &
                                           B.E. in Information Technology           B.E. in Computer Science
                 Communication
          Engineering Society member                    NA                       Engineering Society member


     Limitations:

     •     Simple set of rules comparing education level and work experience
     •     Uniform approach across candidate base
     •     Customary education/work experience
     •     Difficult to differentiate people

20         Footer                                                              Copyright © 2011 Deloitte Development LLC. All rights reserved.
Talent Analytics – Expand the data set

•     How long has the candidate been residing in the city?

•     Does he / she own a house?

•     Is he / she a member of any external agencies / non-profit ventures?

•     What is the commute time to the office?

•     How many promotions has the person had in the last 5 years?

•     What is the average compensation increase that he / she has received in the
      last 3 years?

•     Does the person belong to a Tier I or a Tier II city? (aspirations)


    Predictive models built from these and hundreds of other data elements can better quantify the
                   likelihood and reasoning of future individual employee events.


21    Footer                                                          Copyright © 2011 Deloitte Development LLC. All rights reserved.
Analytics Dashboard – Illustration

                                   Sachin                       Saina                                    Rahul
                            20% less likely than        80% more likely than             30% more likely than
                            average to be a             average to be a                  average to be a
     Likelihood of future
                            successful hire and stay    successful hire and stay         successful hire and
            event
                            with the company for 3      with the company for 3           performance rated
                            years                       years                            above average
                            •   Sub-optimal
                                                        •   Optimal past
                                                                                         •       Sub-optimal
                                employment history                                               employment history
                                                            employment history
                            •   Long Commute –
                                                        •   Short Commute – 1
                                                                                         •       Medium Commute –
       Top 3 reasons            40 miles                                                         15 miles
                                                            mile
                            •   Has been a
                                                        •   Owns a house in the
                                                                                         •       Has been a resident
                                resident of this city                                            of this city for 5
                                                            city
                                for 2 years                                                      years
                                                                                         •       Possible pursuit -
                            •   Unlikely pursuit –
                                                        •   Actively pursue for                  Second tier
      Possible actions                                      national position –                  (possible option for
                                third Tier
                                                            Primary Tier                         local/regional
                                                                                                 position)

If the Predictive Analytics Model is effectively implemented, it allows scarce resources to be better
focused, resulting in measurable benefits.

22   Footer                                                                Copyright © 2011 Deloitte Development LLC. All rights reserved.
Building a Predictive Model
Predicting Attrition – The Holy Grail
Example of potential model variables for active and terminated employees that
may already be available in existing HR systems:

      Employment                Employee               Time and
                                                                              Compensation                        Performance
         Data                  specific data           expense
•    Office Address       •   Home Address        • Hours Worked          •   Salary                         • Performance Rating
•    Department           •   Age                 • Number of Training    •   Bonus                            Previous 5 Years
•    Date of hire         •   Gender                Days                  •   ESOPs                          • Expected promotion
•    Supervisor           •   Education Level     • Vacation/Sick Days    •   Performance Rating               date
•    Supervisor’s         •   Marital Status        Taken                 •   Recognition Awards             • Date of Last
     Performance Rating                                                                                        Promotion
                          •   Number of
                              Dependents                                                                     • Date of 2nd to Last
                                                                                                               Promotion
                                                                                                             • Date of 3rd to Last
                                                                                                               Promotion


Example of potential model variables from external sources:

                                                External Data Elements
• GDP Growth Rate                                            • Average salary increase
• Unemployment Rate                                          • Niche skill vs Easily replaceable skill
• Number of Talent Competitors in the same city              • Additional Macro Economic Variables



24      Footer                                                                        Copyright © 2011 Deloitte Development LLC. All rights reserved.
Produce Attrition-Impact Reports for Employees
                                                                                                    ILLUSTRATIVE

              Sample Model Input                       Sample Model Output
Employee Name                 Saina               Risk Segment                                 High
Employee ID                   1234                Risk of Leaving                              88%
Location/Region               Hyderabad / AP      Actual Cost of Replacement                   $6,000
Employees in Location         6,000               Expected Cost
                                                                                               $84,480
Date of Hire                  01-03-2006          (Risk * Actual Cost)
Rating                        3                   First Most Important Reason                  Time until
                                                  for Risk of Leaving                          promotion
Tenure                        5 years
                                                                                               Supervisor’s past
Base Pay                      $64,000             Second Most Important
                                                                                               retention rate is
                                                  Reason for Risk of Leaving
Position Level                Analyst                                                          low
Working Hours/Yr              1,920               Third Most Important
                                                                                               Long commute
                                                  Reason for Risk of Leaving
Training Weeks/Yr             2
Expected Promotion Year       2011
Vacation Days Taken to Date   16
Commute to Work               > 25 miles


 Actual cost is based on performance rating, level and hiring costs. Reason codes are developed
                    from statistically significant terms in the predictive model.
25   Footer                                                         Copyright © 2011 Deloitte Development LLC. All rights reserved.
Key Roadblocks for Implementing Talent Analytics

           Roadblock                                          Suggested Approach


       Lack of Sponsorship       Gain support from those who derive value from the work being done.


     Unreliable Data Quality &
                                 Ensure data requirements and data integrity are addressed.
            Availability


      Not Aligned to Strategy    Develop metrics from a very clear understanding of the company’s strategy.


      Not Understanding the
                                 Keep in mind the audience and what they might value.
       ―Customer’s‖ Needs


        No Accountability        Set guidelines of expectations and create responsibility among people.


Not Starting Simple and Small    Focus on the ―vital few‖ measures that really make a difference.




26     Footer                                                                       Copyright © 2011 Deloitte Development LLC. All rights reserved.
Points for Discussion

Q1 At the current time,
what are your key needs in                                  Q3 How do you think your
terms of analysis that you                                  needs/wishes around HR
would like to be carried out                                analytics will evolve, when
on your HR data, and what                                   the economy improves?
are the key outcomes and
business decisions that you
are trying to address with     Needs     Tools
that analysis?




Q2 What are your views in
relation to the quality and
                               Issues   Outcomes            Q4 What is your “holy
completeness of the data                                    grail” of HR analytics that
that you have available to                                  you would like to carry
carry out a) basic analysis                                 out, but don’t have the
of your HR data and b)                                      time/data/resources to do
more advanced analytics on                                  so?
your HR data?




27   Footer                                        Copyright © 2011 Deloitte Development LLC. All rights reserved.
Success of Talent Analytics (TA)



                  TA = MC2
                             Measure
                               X
                             Context
                               X
                            Competence



Incisive insights embedded in the right context can drive immense value!


28   Footer                                        Copyright © 2011 Deloitte Development LLC. All rights reserved.
What Are the Key Takeaways?

• Analytics can support and drive change and can be quantified .

• Analytics can assess issues objectively and consistently

• Analytics can add sophistication for HR to manage talent and
  perception.

• Analytics can serve executive leadership in a strategic advisory role

• Analytics can make you look good and reboot your career 




29   Footer                                        Copyright © 2011 Deloitte Development LLC. All rights reserved.
NHRDN Virtual Learning Session on HR Analytics

NHRDN Virtual Learning Session on HR Analytics

  • 1.
    Talent Analytics The elixirto a successful HR career Feb 2012 - NHRD
  • 2.
    Do You…. • Oftencatch yourself saying ―I feel this is the right thing to do‖? • Still use interview conversations to predict the suitability of a potential employee with a reasonable degree of certainty? • Track attrition statistics on percentage points and feel happy if it shows a downward trend? • Have a plethora of training programs but no clear link to business impact? • Want to make an impact but did not know how to do it? …….And, If I wanted to lift my sagging career, what would I do? 1 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 3.
    Moneyball……. • What doesone needs to do when one does not have the money muscle to get talent ? • Has a boss who shows no support for any expense ! • The business and your self respect depends on an outcome that seems challenging to achieve? • And you wish to make a difference !! • THINK DIFFERENTLY and use WHAT is AVAILABLE to you…DATA In God we trust, all others bring Data – Edward Demming 2 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 4.
    Talent analytics –your barometer Workforce planning and deployment Talent management elements Talent sourcing and selection People/leadership development Performance management Rewards and recognition Knowledge management 3 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 5.
    Weather reports –track the talent radar • Those which summarize and compare operational and/or financial data on key workforce variables within defined time frames. – e.g., totals, averages, percentages, and trends. Today • Mathematical models that use multiple internal and external data sources to predict future talent events. • Those which apply one or – e.g., a predictive model two internal data sources to that uses internal and Talent derive useful information. … external employee level analytics data to predict the – e.g., the past education likelihood that a particular experience of job employee will resign in the candidates is compared to next six months and job performance during the supply the reasons for the first year of employment. prediction (e.g., long Maximize commute). performance Yesterday Tomorrow 4 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 6.
    Back to Management reports Talent Analytics Attrition waterfall 2,161 198 9,642 10,000 8,480 801 8,000 6,000 Headcount 4,000 2,000 FY11 Attrition New starts New starts FY12 March attrition Feb 5 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 7.
    Back to Management reports(cont.) Talent Analytics Performance rating movements 1 0 | 0% 2 0 | 0% 3 0 | 0% 0 | 0% 0 | 0% 0 | 0% 0 | 0% Outstanding Very Very good 0 | 0% 0 | 0% good Good 0 | 0% 0 | 0% Good 0 | 0% 0 | 0% Average Average FY09 000 000 000 000 000 000 000 FY08 000 000 000 000 000 000 000 0 | 0% 0 | 0% Avg. Avg. 0 | 0% 6 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 8.
    Weather reports –track the talent radar • Those which summarize and compare operational and/or financial data on key workforce variables within defined time frames. – e.g., totals, averages, percentages, and trends. Today • Mathematical models that use multiple internal and external data sources to predict future talent events. • Those which apply one or – e.g., a predictive model two internal data sources to that uses internal and Talent derive useful information. … external employee level analytics data to predict the – e.g., the past education likelihood that a particular experience of job employee will resign in the candidates is compared to next six months and job performance during the supply the reasons for the first year of employment. prediction (e.g., long Maximize commute). performance Yesterday Tomorrow 7 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 9.
    Back to Employee commitment Talent Analytics Human Resource (HR)/ Culture Employee Switching people practices (Drivers) (Moderators) commitment factors (Goal/Outcome) (Commitment moderators) • Service leadership • Change management • Employee communications Employee Focus • Immediate supervision Employee • Employee growth and development Commitment Switching Costs • Training and education Customer Focus • Affective • Performance evaluation Attachment • Recognition of employee • Willing to be performance Community Focus Proactive Switching Alternatives • Compensation • Intent to Stay • Benefits • Career-life fit Financial Focus • Teamwork/team management • Diversity management • Customer relationship management 8 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 10.
    Weather reports –track the talent radar • Those which summarize and compare operational and/or financial data on key workforce variables within defined time frames. – e.g., totals, averages, percentages, and trends. Today • Mathematical models that use multiple internal and external data sources to predict future talent events. • Those which apply one or – e.g., a predictive model two internal data sources to that uses internal and Talent derive useful information. … external employee level analytics data to predict the – e.g., the past education likelihood that a particular experience of job employee will resign in the candidates is compared to next six months and job performance during the supply the reasons for the first year of employment. prediction (e.g., long Maximize commute). performance Yesterday Tomorrow 9 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 11.
    Back to Retention tracker Talent Analytics Lift chart – demonstrates the effectiveness/benefit of the retention model 40 30 20 Extremely high High attrition risk Relative attrition risk attrition risk 10 segment segment 0 -10 Low attrition risk Moderate attrition risk -20 segment segment -30 -40 1 2 3 4 5 6 7 8 9 10 Deciles Model Equation target = a+b1(tenure)+b2(commute to work)+ b3(pay)+ b4(rating)+b5(training hours)+… This equation is used to give a score to each employee. 10 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 12.
    Back to Retention tracker(cont.) Talent Analytics Solution set to provide an overview of the attrition risk in the organization Organization 4% Attrition risk projections can be analyzed at organizational level Business1 Business 2 Business 3 3% 2% 4% Attrition risk projections can be analyzed at a business level Team 1 Team 2 Team 3 Team 4 1% 2% 4% 5% Attrition risk projections can be analyzed at a team level Attrition risk probabilities are generated for each employee in the organization 11 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 13.
    Weather reports –track the talent radar • Those which summarize and compare operational and/or financial data on key workforce variables within defined time frames. – e.g., totals, averages, percentages, and trends. Today • Mathematical models that use multiple internal and external data sources to predict future talent events. • Those which apply one or – e.g., a predictive model two internal data sources to that uses internal and Talent derive useful information. … external employee level analytics data to predict the – e.g., the past education likelihood that a particular experience of job employee will resign in the candidates is compared to next six months and job performance during the supply the reasons for the first year of employment. prediction (e.g., long Maximize commute). performance Yesterday Tomorrow 12 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 14.
    Talent Analytics –A Model • What is the business issue that we are looking to address? • What kind of data is required to analyze the issue? • How do we look at the data to draw meaningful conclusions? • How are we measuring the effectiveness of these measures? • Can we measure it in $ terms and report to the business? • What insights can we draw from the analysis? • Is it possible to extrapolate past data to predict future outcomes? 13 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 15.
    Analytics - ThePast and The Future
  • 16.
    Evolution of Analytics Predictive Analytics Cross process and functional Analytics Basic Analytics Consolidated reporting Data and Basic Reporting 15 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 17.
    Predictive Modeling Defined Predictivemodeling uses available internal and external data to predict future events at an applicant, employee, or claimant level. Models can be designed to predict a variety of outcomes and have broad based applications. Data Mining — a process, which utilizes a number of mathematical techniques, to analyze large quantities of internal and external data, in order to unlock previously unknown and meaningful business relationships. Predictive Modeling — the application of data mining techniques and algorithms to produce a mathematical model that can effectively predict and segment future events. 16 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 18.
    The 4 CriticalModeling Questions 1. Is there a compelling problem or opportunity? • What is the business case scope and size? • What are the costs, including opportunity costs? 2. Do we have the data we need? • What is the state of available data and our ability to access it? 3. Can we segment or predict potential outcomes and does it put us in a position to make a difference? • Is there a basis to build a predictive model? • Will the model output help us solve the defined business problem? 4. Can we effectively act upon the predictive model output? • What is our change readiness and ability to implement? • Will implementation drive the required ROI? • What is our political and legal climate? 17 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 19.
    What Can PredictiveModeling Do? • Predictive models use available internal and external data to predict the likelihood of future events at the customer or employee level • Models are deployed by businesses to direct limited resources to the actions that will yield the largest economic benefit For example… Similarly… A business unit uses predictive HR can use predictive models to models to maximize ‘customer maximize ‘employee lifetime lifetime value’ value’ • Targeting new customers • Recruiting and hiring new employees • Optimizing pricing, customer service and costs • Optimizing development and performance • Focused customer retention • Focused employee retention 18 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 20.
    Talent Acquisition –A Case in Point Problem Statement: How do we review more qualified candidates, faster, with improved accuracy, and with less cost? 150,000 applications p.a. which results in an eligible candidate pool of 81000 Initial screening brings this number down to 8000 Further filtration occurs through multiple rounds of interviews 2100 Offers / 1675 accepted 19 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 21.
    Traditional Recruiting Data •Who would be the most successful and Who would be the long term employee? Sachin Saina Rahul 10 years of work experience 10 years of work experience 10 years of work experience 4 previous employers in past 10 2 previous employers in past 10 1 previous employer in the past years years 10 years Current employer is large Current company is small Current company is a microchip technology company technology company mfg. company Attended Tier I Engineering Attended Tier II Engineering Attended Tier III Engineering College College College B.E. in Electronics & B.E. in Information Technology B.E. in Computer Science Communication Engineering Society member NA Engineering Society member Limitations: • Simple set of rules comparing education level and work experience • Uniform approach across candidate base • Customary education/work experience • Difficult to differentiate people 20 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 22.
    Talent Analytics –Expand the data set • How long has the candidate been residing in the city? • Does he / she own a house? • Is he / she a member of any external agencies / non-profit ventures? • What is the commute time to the office? • How many promotions has the person had in the last 5 years? • What is the average compensation increase that he / she has received in the last 3 years? • Does the person belong to a Tier I or a Tier II city? (aspirations) Predictive models built from these and hundreds of other data elements can better quantify the likelihood and reasoning of future individual employee events. 21 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 23.
    Analytics Dashboard –Illustration Sachin Saina Rahul 20% less likely than 80% more likely than 30% more likely than average to be a average to be a average to be a Likelihood of future successful hire and stay successful hire and stay successful hire and event with the company for 3 with the company for 3 performance rated years years above average • Sub-optimal • Optimal past • Sub-optimal employment history employment history employment history • Long Commute – • Short Commute – 1 • Medium Commute – Top 3 reasons 40 miles 15 miles mile • Has been a • Owns a house in the • Has been a resident resident of this city of this city for 5 city for 2 years years • Possible pursuit - • Unlikely pursuit – • Actively pursue for Second tier Possible actions national position – (possible option for third Tier Primary Tier local/regional position) If the Predictive Analytics Model is effectively implemented, it allows scarce resources to be better focused, resulting in measurable benefits. 22 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 24.
  • 25.
    Predicting Attrition –The Holy Grail Example of potential model variables for active and terminated employees that may already be available in existing HR systems: Employment Employee Time and Compensation Performance Data specific data expense • Office Address • Home Address • Hours Worked • Salary • Performance Rating • Department • Age • Number of Training • Bonus Previous 5 Years • Date of hire • Gender Days • ESOPs • Expected promotion • Supervisor • Education Level • Vacation/Sick Days • Performance Rating date • Supervisor’s • Marital Status Taken • Recognition Awards • Date of Last Performance Rating Promotion • Number of Dependents • Date of 2nd to Last Promotion • Date of 3rd to Last Promotion Example of potential model variables from external sources: External Data Elements • GDP Growth Rate • Average salary increase • Unemployment Rate • Niche skill vs Easily replaceable skill • Number of Talent Competitors in the same city • Additional Macro Economic Variables 24 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 26.
    Produce Attrition-Impact Reportsfor Employees ILLUSTRATIVE Sample Model Input Sample Model Output Employee Name Saina Risk Segment High Employee ID 1234 Risk of Leaving 88% Location/Region Hyderabad / AP Actual Cost of Replacement $6,000 Employees in Location 6,000 Expected Cost $84,480 Date of Hire 01-03-2006 (Risk * Actual Cost) Rating 3 First Most Important Reason Time until for Risk of Leaving promotion Tenure 5 years Supervisor’s past Base Pay $64,000 Second Most Important retention rate is Reason for Risk of Leaving Position Level Analyst low Working Hours/Yr 1,920 Third Most Important Long commute Reason for Risk of Leaving Training Weeks/Yr 2 Expected Promotion Year 2011 Vacation Days Taken to Date 16 Commute to Work > 25 miles Actual cost is based on performance rating, level and hiring costs. Reason codes are developed from statistically significant terms in the predictive model. 25 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 27.
    Key Roadblocks forImplementing Talent Analytics Roadblock Suggested Approach Lack of Sponsorship Gain support from those who derive value from the work being done. Unreliable Data Quality & Ensure data requirements and data integrity are addressed. Availability Not Aligned to Strategy Develop metrics from a very clear understanding of the company’s strategy. Not Understanding the Keep in mind the audience and what they might value. ―Customer’s‖ Needs No Accountability Set guidelines of expectations and create responsibility among people. Not Starting Simple and Small Focus on the ―vital few‖ measures that really make a difference. 26 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 28.
    Points for Discussion Q1At the current time, what are your key needs in Q3 How do you think your terms of analysis that you needs/wishes around HR would like to be carried out analytics will evolve, when on your HR data, and what the economy improves? are the key outcomes and business decisions that you are trying to address with Needs Tools that analysis? Q2 What are your views in relation to the quality and Issues Outcomes Q4 What is your “holy completeness of the data grail” of HR analytics that that you have available to you would like to carry carry out a) basic analysis out, but don’t have the of your HR data and b) time/data/resources to do more advanced analytics on so? your HR data? 27 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 29.
    Success of TalentAnalytics (TA) TA = MC2 Measure X Context X Competence Incisive insights embedded in the right context can drive immense value! 28 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.
  • 30.
    What Are theKey Takeaways? • Analytics can support and drive change and can be quantified . • Analytics can assess issues objectively and consistently • Analytics can add sophistication for HR to manage talent and perception. • Analytics can serve executive leadership in a strategic advisory role • Analytics can make you look good and reboot your career  29 Footer Copyright © 2011 Deloitte Development LLC. All rights reserved.

Editor's Notes

  • #3 So, I looked at some of the things that I faced as an HR professional….
  • #4  I love sports and am always amazed by what happens all around and the way athletes and sportsmen grow in their lives. The IPL gave way to a greater fantasy of watching teams come together and play. In 2007, the Rajasthan Royals won the IPL. They were the unlikely winners by a long shot. There were no million dollar pinch hitters, they had an old Shane Warne who was not known to be an inspiring leader. Not sure if Shilpa Shetty had a role but, it just seemed impossible for them to win. A couple of years later I read a book by Michael Lewis called ‘Moneyball’. It was about a baseball team- Oakland Athletics, who won 20 games on trot, an unlikely event, almost impossible. But it happened. Owed largely to a Manager Billy Beane ( played by Brad Pitt). Billy was pushed to the corner because his owner did not buy expensive players, his team had lost some players to other teams and they were left with a bunch of not so great guys. Beane is committed to win and has an ambition to beat the best, in an economical way. He had to make some drastic change to his strategy. He had met a Yale grad, Peter Brand, who based on some statistics, said that there were some fantastic players available. Billy went through reams of data, agreed with his young partner and made the decision. Together they recruit a bunch of undervalued players, misfit boys, who have the potential to be match winners. Billy had to take those chances. Translating theory into practice is not easy. He did and the rest is history. Rajasthan Royals perhaps had such a story, except it was never told this way!!This story has relevance to our current state: big plans, small purse, big ambition, no money. But we have to win. No Brand, small brand, have to recruit, and the best, and win the war for talent. There are many areas where this plays out. Where do we recruit? Who do we retain? Do we have to go after all the big boys? And all of this in the most economical way!!. Organizations spend millions of dollars to make business work, and yet Talent remains their number one concern area. This webinar is about to let you know what you can do to make your CEO listen to you and get it right. Beane’s problem was that wealthier teams such as the New York Yankees,With many multiples of the A’s salary Teams such as the New York Yankees, with many multiples of the A’s salary budget, could out-bid the A’s when scouting for new talent. Beane addressed this problem with crucial insight: baseball scouts often use flawed reasoning and fallible “gut feelings” or “professional judgment” when selecting baseball players. FROM MONEYBALL TO WORKFORCE INTELLIGENCEThe story of how Billy Beane used analytics to identify undervalued baseball players has far-reaching implications for many industries. The most obvious parallel is the war for talent. Baseball is not the only domain where the stakes are high when it comes to attracting and retaining talented employees. Consider the following facts: Most companies must devote anywhere between 40 and 70 percent of their operating expenses to compensation, benefits and other employee-related expenses.8 In many domains, a rule of thumb estimate of the cost of replacing an employee is 1.5 times that employee’s salary. Finally, the business press is replete with warnings that as the population ages, the competition to attract and retaintalented workers will intensify. Yet most large organizations still make their hiring decisions using a highly labor-intensive and subjective approach often centering on subjective evaluations of candidates’ performances at interviews.
  • #14 So, I wish to speak to the fewmost important things here:First things first:We must know what data we have that we report out.Do we draw insights from it? This is something that is key to your career…INSIGHTS.Do you have a problem statement that can be spelt out?In short:A clear understanding of the problem at hand.The ability to speak to the Business and let them know the value once there is a problem that has been solved.The greatest value of Analytics is the ability to predict the future…that is INSIGHT.Eg. I had a salary issue on hand…..and the way I asked for a 12 % hike was not what everyone was doing…..that would be blue murder !! So I erad up a lot , looked at history, drew a model of the past, and predicted that we would more likely end up at an x% in the next few years. I spoke to the CFO of the company and had my data pat on….at the end of 30 mts, the CFO was so pleased that he wondered why I didn’t ask for more, and I said that I had a Business to protect  it drew a loud laugh in him.So, let us speak of how analytics has evolved:
  • #16 This is where it all begins : if we have Time on one axis and Value on the other, this is how it would look. We start with the basics reporting. I remember I had to report out each day of the number of workers that were in the factory, and the overall numbers that we had as our headcount. Later, I gave that reprot as a Monthly report. It occurred to me that there was a pattern of those present and it led to some analysis. And if I were to look at other plants / depts around, it would be cross functional. The final one is when I am able to predict the attendance on any particular day of the year, given the data that I have. Aah !! Now I had my bosses attention.
  • #17 There is a new buzz word in the market : Context Intelligence !! It means that if we have all the parameters and build a heuristic model around it, it may well be that we could predict outcomes. Eg. How many of those offered would actually turn up / do well etc.Where can we use this intelliegence? Recruitment, Talent Analytics and Planning, Resource Allocation, Career Planning, Leadership pipeline, Engagement and Retention, Compensation etc.And this is not impossible, as I will show you later in the presentation.
  • #19 I had to make a presentation to my CFO on whether the salaries in India would remain competitive for the next few years viz a vis the US. To me this was a great opportunity to show that even though we had an increase of about 10 to 12 % hikes, India would remain competitive for us, and that the actual cagr on compensation would be 4 % !!