Lyndon Sundmark: People Analytics - A way to do HR?
People Analytics Conference
Website - https://pacamp.org/
Youtube - https://www.youtube.com/channel/UCeHtPZ_ZLZ-nHFMUCXY81RQ/featured
FB - https://www.facebook.com/pacamporg
2. My Background and
Contact Information
• Background
• 40 years experience with integrating HR, IT, and Statistical
Analysis skills into HR work. This also includes HR software
development and the development of an HR Data
Warehouse
• Retired – but continue to write in this field and informally
mentor periodically
• MBA 1980 (major – HR related)
• Author of book "Doing HR Analytics - A Practitioner's Handbook
With R Examples":
• https://www.amazon.com/Doing-HR-Analytics-
Practitioners-
Handbook/dp/1973716372/ref=sr_1_1?ie=UTF8&
qid=1511821165&sr=8-
1&keywords=doing+hr+analytics+-
+a+practitioner%27s+handbook+with+r+examples
• Contact Information
• Email: lvsund@outlook.com
• LinkedIn: https://www.linkedin.com/in/lyndon-sundmark-
mba-59272a/
3. My Presentation Outline
1. What is wrong with how
People Analytics is currently
defined?
2. Defining it to maximize its
potential contribution
3. Giving It A Broad Scope
4. How that scope fits into
the levels of analytics
5. A practical example of
machine learning in People
Analytics
6. How do we encourage HR
professionals to see People
Analytics as a way of doing
HR?
7.What is the challenge to
HR professionals?
8. What is the challenge for
HR degree, diploma,
certification providers?
4. 1. What is
Wrong With
How People
Analytics Is
Currently
Defined?
Too Many Definitions
Plethora of definitions
causes:
Its important to have
definition that enables us as
HR to move this discipline
forward and to get actual
organizational work done
•Vendor definitions
•Recruiting definitions
•Literature definitions
•Social Media definitions
•lack of standardization
•confusion
•unnecessary organizational
territorial disputes
5. 2. Defining it
to maximize
its potential
contribution
People Analytics
• “data-driven”, ”evidence based” HR
Management and Decision Making
• Implications
• Ultimately – its intent is not as an
add-on to HR
• It’s a way of ‘doing’ HR.
• Million $ question- what drives our
HR Management and Decision
Making choices currently?
6. 3. Giving It A Broad Scope
• People Analytics needs to include:
• Traditional HR Metrics & Associated Enabling Technologies
(providing a ‘people’ picture to the organization)
• Includes HR measurements, KPIs, Dashboards, Data
Warehousing of HR Data, Slice And Dice Tools etc.
• Ongoing monitoring and improvement in HR Services productivity-
TQM and QI
• Continuous evaluation and improvement in HR practices and
methodologies and their outcomes through the application of
machine learning and AI
• This is what enables HR Management and Decision-
Making to be ‘data-driven’ and ‘evidence based’
• The reason why do any of these things is improvement of
organization performance through HR practices and better
HR decisions
7. 4. How that scope fits into the levels of analytics
Descriptive Diagnostic Predictive Prescriptive
-HR Graphical
Management Reports
-HR Metrics
-HR KPIs
-Dashboarding
-Storytelling
-Data Warehousing
-Slice and Dice Tools
-Statistical Tools and
Analysis
Univariate
Multivariate
- HR Services
Operational
Improvement- TQM
Machine Learning
and AI
- HR practices and
methodology
improvement
Going from
predicting what an
outcome will be to
engineering a desired
outcome?
-sensitivity analysis of
ML models
-tweaking the models
8. 5. A practical
example of
machine
learning in
People
Analytics
• Employee Attrition – predicting those at possible
risk of leaving
1. Thinking informationally and obtaining data
2. Exploratory Data Analysis
3. Applying selection of machine learning
algorithms(models) to the data
4. Evaluating the algorithms(models)
5. Using best models to predict on new data
• This type of activity can be done with FREE tools
• R
• Python
9. 6. How do we encourage more ‘evidence
based’ and ‘data driven’ HR?
• Encourage the development of HR Metrics and data
warehousing.
• Encourage the use of ‘ticketing’ or ‘service request
tracking’ systems
• Encourage the use of machine learning and AI
wherever possible
• Systematically evaluate each HR functional area and brainstorm
answers to following questions:
• What are the outcomes that we are trying to encourage or
discourage in this area of HR?
• Are those outcomes categorical events or statuses, or are
the outcomes an amount we are trying to predict ahead of
time?
• What machine learning algorithms are designed for
predicting categorical outcomes versus amounts?
• What data do we have in this HR functional area that is
associated with these outcomes and where is this data?
10. 7.What is the
challenge to
HR
professionals?
We no longer have the luxury of being trained up just in
the traditional disciplines of HR and in the traditional way
HR must also be trained up to understand HR
“informationally”
• Includes awareness that all of HR activities typically generate
information, but not all of that information is currently recorded and
stored.
• Includes data retrieval skills, knowledge of the basics of data
warehousing and why its needed, and such concepts as late arriving
data and its relation to HR Metrics
HR must also be trained up in Statistical Analysis, Research
Methods, and Machine Learning/AI
• R
• Python
11. What is the challenge for HR
degree, diploma,
certification providers?
• Bottom line
• HR curriculums and certifications need to include
courses that go beyond the traditional HR functional
domain areas.
• Courses need to be provided that are IT related that
teach data retrieval skills, thinking informationally
with respect to HR, and differences between HR
information systems and data warehouses and why
data warehouses are important. Preferably- learning
with HR data examples
• Statistics courses that provide a fairly robust level of
knowledge and understanding of univariate and
multivariate statistics, and machine learning and AI –
WITH HR RELEVANT EXAMPLES.
Editor's Notes
Thank you for the opportunity to do this conference presentation
My name is Lyndon Sundmark
The focus of my presentation is to ask and consider a question- Is People Analytics a way to do HR?
Before I get into my presentation- just a little bit about me- because a lot of what I will present is a product of my background.
I have had the privilege in my career of both using the building blocks on modern day People Analytics and also see the evolution of the tools and building blocks over the same amount of time
Yes- the building blocks have been around for that long. People Analytics is the recognition of those building blocks
In order to answer the question that this presentation is focused on- I would like to cover the following areas
You need this broad a scope or vision in People Analytics to prevent it from becoming just a project, initiative, or a fad.
In this vision - HR is not ‘business as usual’