Analytics Driving Action - Building a Data-Driven HR Function
1. Analytics Driving Action
Jonathan Sidhu
Executive Program Manager
Business Analytics Transformation
@jmsidhu
Mark Tristam Lawrence
Learning Intelligence Leader
Global Business Services
@mtlawrence
#PADDHR
2.
3. Agenda
IBM – Who we are and what we do
Workforce Analytics - Background
Analytics in Action
Getting Started
4. IBM - a virtual social community
72% of us outside Americas
64% workforce in Services business
55% workforce has less than 5 years
service
36% of employees work remotely
12% from acquisitions & outsourcing deals
1% on global assignments
5. Agenda
IBM – Who we are and what we do
Workforce Analytics - Background
Analytics in Action
Getting Started
6. Source 1: 2012 IBM CEO study: Q24 ―What do you see as the key sources of sustained economic value in your organization?‖
Source 2: SHRM Human Capital Benchmarking Database, 2011
Products / services innovation
Human capital
Customer relationships
Brand(s)
Business model innovation
Technology
71%
66%
52%
43%
33%
30%
Human capital is the leading cited source of economic value...
...but, CEOs face significant workforce challenges.
The average turnover in the
U.S. is 15% per fiscal year.2
Total costs of replacement can reach
200% of an employee‘s annual salary.2
Key sources of sustained economic value1
CEO Study
7. HR Value
StrategicSmarter WorkforceOperational
Smarter HR Operations
Succession
Management
Learning &
DevelopmentCompensation
Management
Performance
Review
Performance
Planning
Talent
Acquisition
Workforce
Planning
Absence
Management
Payroll
Benefits
Time &
Attendance
Scheduling
& Staffing
Analytics are critical to HR operations and workforce
effectiveness
8. We redefined IBM‘s HR Strategy
Thought
Leadership Develop deep expertise in business, HR,
execution methods and systems thinking
to enable IBM‘s ability to produce value for
clients, IBMers, investors and communities
Analytics
Use technologies to capture, analyze and
integrate data that will yield rich insights
and advance the science of predicting,
shaping and adapting to business trends
Collaboration
Use contemporary technologies and
techniques within the HR community to
enable seamless global communication to
develop, test, refine and implement the
best ideas
9. Time and Resources
ValueandImpact
Based on : Competing on Analytics, Davenport and Harris
Data Management
Consolidation of data
Data quality and accuracy
Basic Reporting
Standard reporting that is reasonably automated
‗Slice and dice‘ data based on standard variables
Benchmarking
Key Performance Indicators (KPIs)
Performance Measured against Best Practices
Analysis
Multi-dimensional analysis to better
understand business challenges
Advanced Analytics
Segmentation, Predictive
Modeling & Optimization
Efficiency
Effectiveness
Business Impact
Business Analytics Model
10. Enterprise-wideReporting
PredictiveAnalytics
InternalSurveys
ExternalAnalytics
Objectives
• Integrate BI into HR as an ‗everyday‘ tool
• Develop deeper analytic and predictive modelling
skills
• Develop key enterprise-wide reports and
scorecards
• Provide better insights from analytics to inform
strategy
• Grow analytic skills in emerging countries
Purpose: To embed a culture of analytics
within the HR organization
SocialAnalytics
IBM Workforce Analytics - a clear purpose…
11. Agenda
IBM – Who we are and what we do
Workforce Analytics - Background
Analytics in Action
Getting Started
12. Can attrition be predicted?
Questions
Can we highlight individuals who
have the highest propensity to leave
an organization?
Can increased compensation reduce
the rate of unwanted attrition?
Method
Analysis of 5 years longitudinal data
Compensation & attrition analysis
based on historical data
Visualization – use of heat maps
Geography
Brand
Other
Type of
Hire
Education
Type
Segment and sub segment of population
13. Can we see attrition risks & reduce unwanted attrition?
Retention Case Selection
Action Optimization—Identify retention
cases and targeted actions to retain
them
Attrition
Which employees are most likely
to leave? What kind of actions,
programs and investments will
reduce attrition in the most
effective way?
How likely is each person to leave, and why?
Attrition Hot Spots
Identify high-attrition clusters
Derive attrition ―rules‖
Estimate FUTURE attrition
Understand response to incentives
How well-connected are those
employees most likely to leave?
What actions will yield the
best outcomes?
Benefit($k)
Cumulative Net Benefit is maximized at $9M…
... yielding an attrition reduction of 2.7 pts
14. Our results?
Attrition can be predicted
We can highlight those who have the highest propensity to leave
the organization
Proactive retention efforts can reduce rates of unwanted attrition
15. Can we measure Employment Risk?
Problem Statement
Can collective employment risk be quantified?
Can a ‗risk index‘ be built?
Methodology
Statistical analysis of 5 year longitudinal internal and external data
Tens of metrics analyzed for stability to find a suitable set of dependent
metrics
Index stabilized over time
Visualization – use of heat maps, indices and maps
Built into BI tools
16. Predictive Analytics – Headlights for Employment Risk
Quality of Life Index
Cultural Values
(e.g. Hofstede)
Span of control
Employee Complaints
Inflation & Unemployment
Surveys
Headcount & Voluntary Attrition
Employment
Regulation
Tenure, Band, Rating
Political Instability
Country Level Volatility
Data Is Simulated
17. Our results?
Employment risk can be quantified
Usable risk indices can be built to:
Inform & coach
Assist leaders in managing overall
enterprise risk posture
Provide insight to HR leaders
18. Smarter Learning Analytics
Problem Statement
Education performance is judged on volumes, not impact or value
Learning organisations struggle to align with rapidly changing business models
Methodology
Back to basics – why do we provide learning?
Formulation of new metrics, dependencies and relationships
Identification, acquisition and modelling of new datasets
Creation of new reports with optimal visualisation
Standardisation of formatting, branding, platform and access
Democratisation of data!
19. Smarter Learning Analytics – Evaluation
Learner Reaction (Level 1) and
Skill Usage (Level 3)
Resource: Free up dedicated
FTE and increase self-service
Speed: Reduction of time
from training to reporting
Impact: Increase in available
analytics; more powerful
communication
20. Smarter Learning Analytics – ROI
Return on Investment (Level 4)
Automate: Reduce emails from
learning organisation and enable
learners to see value
Standardise: Common
methodology to provide scalable
solution
Impact: Draw direct line of sight
between learning and revenue
21. Smarter Learning Analytics – Alignment
Business Impact (Level 5*)
Delivery Excellence:
Identify and reduce factors attributed to learning or skills in troubled projects
Practitioner Utilisation:
Challenge perception that short-term time away from the client harms longer-
term practitioner utilisation targets
Grow Talent:
Demonstrate the impact of learning in promotion, progression, attrition and
retention
22. Smarter Learning Analytics - Findings
Can we demonstrate the value of Learning?
Can we align metrics with current business decisions?
In practice:
Which learning activities do we prioritise; and which need revision?
Where are we effectively addressing skill gaps; and where aren‘t we?
Why should we invest in learning; and should we increase investment?
23. Can we leverage Social ‗Big Data‘ productively?
Problem Statement
Can we see what employees are saying on social media about our
company?
Can we get ‗real time‘ feedback on our business from our own people?
Methodology
Consumer insight tool redesigned for employee insight
HRIS combined with social media
Highly visual design
‗Opt in‘ consent
24. Analyzing ‗big data‘ created by social interactions
→ Determine unexpected
affinities across multiple
analytic dimensions
→ Discover related topics
above and beyond our
initial search
→ Glean employee
sentiment across
company & by segment
→ Perform trend analysis of
sentiment over time
25. The Key Advantages of Social Analytics
Hear ‗pulse‘ of the organization
Augment surveys with better understanding
of what employees think
Tailor services & programs
Ability to act in nearer real-time
Leverage existing social media footprint
Understand what competitors‘ employees
are saying about their employers
26. Our results?
We can see (and use) what employees are saying on social media
about our company
We can get ‗real time‘ feedback on our business from our own people
that provides insight and allows executives to take action
27. Agenda
IBM – Who we are and what we do
Workforce Analytics - Background
Analytics in Action
Getting Started
28. Analytics: The New Path to Value: Links to the full study, a 22 minute video and presentation highlighting the key findings (https://ibm.biz/newpathvalue)
IBM Institute for Business Value +
Surveyed 3,000 executives, managers
and analysts plus extensive interviews
Respondents represent more than 30
industries in 108 countries
Interviews with IBM and MIT thought
leaders
Analysis by IBM and MIT teams
29. Organizational obstacles, not data or financial
concerns, are holding back analytics adoption
Ability to get the data
Lack of management bandwidth due to competing
priorities
Lack of skills internally in the line of business
Lack of understanding how to use analytics to improve
the business
Culture does not encourage sharing information
Ownership of the data is unclear or governance is
ineffective
Lack of executive sponsorship
Concerns with the data
Perceived costs outweigh the projected benefits
No case for change
38%
34%
28%
24%
23%
23%
22%
21%
21%
15%
Primary obstacles to widespread analytics adoption
Organizational
Data
Financial
Source: Analytics: The New Path to Value; https://ibm.biz/newpathvalue
30. Lesson Learned
• Cultural shift from data extraction to business analytics
• Stakeholders need to set clear priorities for data and analytics
• Need a broad range of skills
• Simplicity and elegance outweigh ―bells and whistles‖
• Take some risks with new tools – open people‘s minds to the
opportunity
• If you do what you‘ve always done, you‘ll get what you‘ve always got
33. Additional IBM Institute for Business Value White Papers
a) Business Analytics & Optimization for the
Intelligent Enterprise (2009)
b) Breaking away with Business Analytics &
Optimization (2009)
c) Analytics: The new path to value (2010)
d) Analytics: The widening divide (2011)
e) Analytics: The real world use of big data (2012)
f) Analytics: A blueprint for value (2013)