Talent Analytics and Big Data
Talent Analytics and Big Data
Talent Analytics and Big Data
Research report
November 2013
WORK
WORKFORCE
WORKPLACE
WORK
WORKFORCE
WORKPLACE
Our focus on the workplace includes how organisations are
evolving and adapting, understanding of culture, trust and
engagement, and how people are best organised, developed,
managed, motivated and rewarded to perform at their best.
About us
The CIPD is the professional body for HR and people development. We have over 130,000 members internationally
working in HR, learning and development, people management and consulting across private businesses and
organisations in the public and voluntary sectors. We are an independent and not-for-profit organisation, guided in
our work by the evidence and the front-line experience of our members.
Contents
Executive summary
Introduction 3
1 Silos: whats getting in the way?
10
13
16
21
Conclusion 24
References 25
Acknowledgements
The CIPD wishes to thank all organisations that participated in our research, both indirectly through providing us
with data from our HR and L&D surveys and directly from the organisations we were able to profile. Those who
attended our engagement events including our July forum event and our Google hangout on September 20
also contributed, as did participants at the City HR forum. Thanks are due to Oracle especially Andy Campbell,
HCM Strategy Director, and his marketing colleague Saurabh Rastogi for sponsoring this key area of research.
Wed also like to thank the various thought leaders who have shared their insight.
Executive summary
Talent analytics and big data are
now must-have capabilities in HR.
As the business world is transformed
by the sheer volume, speed and
availability of data, and as the search
for competitive advantage intensifies,
data about people and performance
becomes ever more critical. Much
has been written on the issue of
talent analytics, but most of this
debate has focused around three key
dimensions:
Technology
What systems, processes and
infrastructure drive data and talent
analytics? What platforms are
being used and how can we use
systems such as Oracle, Hadoop
and the like to develop a coherent
data strategy?
Silos
Silos are the structural and systems
obstacles to timely and efficient
access to data and the ability to
use and share it. This is caused by
a combination of structural barriers
within both HR and the business
which impede the sharing of data.
The system silos are those pertaining
to infrastructure around data
analysis with issues such as systems
incompatibility, security and hosting
concerns, and IT skills issues.
Techniques
How should we develop an approach
to talent analytics? How should we
define, store and share data? How
should it be analysed? This aspect
covers everything from defining
employee turnover to predicting
patterns of employee behaviour.
Talent
Who should deliver our analytics
capability? Should we develop new
talent pools and, if so, what types?
How do we recruit the scarce
talent to resource this growing
capability need?
Solutions
Strategic
Our survey research shows that
these are considerable challenges,
but practice we discovered among
organisations undertaking the
challenge shows that there are
solutions. These are based on
developing a strategic and tactical
mindset which:
Introduction
You wont go far these days
without someone analysing your
data. Whether its a supermarket
using your passion for chocolate
or driving magazines to make
sure you pick one up on your
next order or a council wanting
to improve its understanding of
its citizen base through surveys,
everyone wants to analyse you.
Many of the organisations we work
in operate on the principles of
advanced analytics. For example,
banks generate a whole array of
data about where people live,
their credit history and spending
patterns, as well as lifestyle
decisions such as birth, marriage
and death. Similarly, companies
in the retail sector use predictive
analytics to estimate how to stock
up on sausages for a barbecue
weekend and how to upskill their
staff to sell the latest smartphone.
The idea of data as a business
prediction tool is not new or novel,
but the intensity and sophistication
with which its now being used is
quite new. Business departments
such as marketing and logistics
and specialised agencies operating
in many different sectors from
healthcare to education pore over
Big data is
essential to HR
and L&D because
it allows the
conversations
and connections
which have
tended to be in
the realm of the
immeasurable to
be captured and
leveraged.
Variety
Velocity
Volume
...there is still a
significant gap
in our ability as
HR professionals
to be datadriven and
evidence-based
in our decisions.
Figure 2: HR draws insight from data to stimulate change and improvement in the
organisation (% agreeing)
HR leaders
63%
Business leaders
21%
Many HR people
are reluctant
to engage with
numbers or are
passive towards
them.
60
50
48
40
30
29
23
20
10
Agree
Neither
Disagree
The promise of
talent analytics
and big data
is that they
almost certainly
will move HR
forward in terms
of analysis and
insight.
Silos
Skills
Suspicion
programmes to talent
management and key projects
and assignments
performance data
performance ratings and data
captured from the use of
instruments such as 360, goal
attainment, talent, succession
programmes and talent and
assessment.
The problem is that this data is
often diffuse and difficult to access.
This is often because of what we
term silos. Figure 5 highlights the
two different types of silos that
exist: structures and systems.
Structural silos
The first of these silos are
structural: in other words,
Structures
Systems
HR compartments
Incompatible technology
Permission problems
IT skills issues
Legacy systems
Within HR teams
there is often a
solid separation
between HR
functions and a
lack of productive
collaboration.
Table 1: Difficulties encountered in analysing learning and development data (by organisation size) (%)
Small and medium
<50 and 50249
Medium/
large 250999
Large/very large
1,0004,999 and
>5,000
45/61
68
67/69
43/57
52
66/65
46/53
45
37/37
23/27
28
37/44
34/32
31
30/37
22/24
36
34/36
Source: CIPD Learning and Talent Development survey 2013 (sample size: 1,004; base: 742)
Systems silos
The other type of silo can be
described as systems silos in
other words, having systems
which are incompatible and poorly
integrated. Many organisations
operate with either hybrid data
systems or have linked systems
together inappropriately. This can
lead to systems which cant talk to
each other.
Separately, concerns about data
security can also cause silos.
Sometimes the smooth flow of
access is impeded by permission
problems which plague everything
from social media access to critical
information. Its understandable
in a business world nervous about
hacking and security concerns, but
sometimes these obstacles can stifle
the collaborative approaches needed.
Generally for talent analytics to take
off and for big data to be harvested,
the approach needs robust but
permissive security. Restricting access
to just a few limits insight. Solutions
already exist to help overcome these
problems in many of the integrated
human capital management
programmes available.
IT and database skills issues can
also get in the way, with the ability
to use database query languages
and an ability to programme often
prerequisite to run some fairly basic
data enquiries. This can provide
skills problems often requiring the
resourcing of extra capability. Linked
to this are the legacy systems issues
and the systems transition problems
evident when organisations
inherit older systems and need to
integrate and use them alongside
newer technology. Coupled with
that challenge is the transition
from one system to another, such
as the migration from SAS to
Oracle and vice versa. Many of our
organisations were going through
this difficult migration challenge as
we conducted our research.
...there is a
long-standing
perception
both within
and outside of
HR that people
within the
function do not
have enough of
the analytical
skills needed
to develop
a coherent
and credible
approach to
analytics.
Figure 6: Skills and smarts in talent analytics and big data for HR
working on understanding
issues such as segmentation and
optimisation of talent. This requires
deep modelling and statistical tools.
Finally, the pinnacle of capability is
defined as the advanced analytics
and predictive skills the ability
to model scenarios, account for
and forecast risks and design
predictive programmes and models
such as algorithms.
The issue of
big data and
HR has caught
fire recently
with a welter of
reports, books and
commentary.
Variety
Velocity
Volume
Figure 7: Gartners three Vs of big data derived from McAfee and Brynjolffson
More HR business
by volume and value
being done on the
Internet.
Information available
everywhere and
anywhere, and
increasingly to
everyone.
Fast, efficient, cheap
and multi-channel
means more HR can
be generated and
captured.
Greater range
of devices and
platforms.
Social media
capturing and
transmitting data.
More aspects of
life, networked
smartphones to
embedded chips,
can be used in
workplace settings.
Value
Increasing speed,
scale and scope of HR
information.
Variety
Velocity
Volume
Integrate systems
and structures.
Build skills and
smarts for analysis.
Encourage
collaboration and
connectivity around
data and information.
Use data as a
transformational tool.
Figure 9: Suspicion and scepticism in HRs approach to talent analytics and big data
Biases, beliefs,
behaviours
H
R data has been
backward-looking and
insular.
M
any have preference for
big picture and ambiguity
over analysis.
C
onflicting demands
and expectations of
organisations.
Fears
D
ata reduces and
dehumanises people to
units of analysis.
Expectations treadmill.
D
ependency on external
capability means HR could
be left behind in the skills
race.
Even in cases
where business
leaders havent
required HR to
be analytical and
evidence-based,
that will certainly
change.
Figure 10: Priority objectives for HR leaders with their teams (%)
59
55
52
51
49
49
45
41
40
40
39
36
35
Dont know
Not applicable no identified key priority areas for
building HR capability
3
2
Other 1
Source: CIPD HR Outlook, autumn 2013: a variety of leader perspectives.
Transformational approaches
Make analytics a continuous
transformational project
If people analytics and datadriven approaches are to take
off, they should be driven as
transformational projects with full
executive team ownership and
enthusiastic and driven day-to-day
management. The transformation
is likely to be incremental and
celebrated by the victory of small
wins and continuous review. Where
were we? Where have we got
to? Where do we go next? The
If people analytics
and data-driven
approaches are
to take off, they
should be driven as
transformational
projects with full
executive team
ownership.
Figure 11: Strategies and solutions for managing talent analytics and big data in HR
Tactical
Transformational
Make analytics
a continuous
transformation project.
Focus it on key business
priorities.
Move it up the HR
capability agenda.
Collaborate
and connect
through data.
Conclusion
When we started this research
we scoped our research questions
to take in the emerging agenda
for HR of engaging with talent
analytics and big data. What we
found is that the pursuit of HR
capability in both of these areas is
at a fairly basic stage. We asked
practitioners to identify themselves
as aspiring-edge or leading-edge
in their approach towards people
analytics issues. We wanted to
deliberately sense-check where
practitioners place themselves on
the analytics spectrum. Most are
happy to be seen as aspiring, but
the surprise is that even those that
we consider to be at an advanced
level of capability are reluctant
to say they are pushing the
boundaries.
Most are very much defining an
approach towards data or, to be
more precise, metrics. They are
ensuring that the numbers add
up and have credibility. They want
their key HR data to be reliable and
consistent on key business driver/
cost issues such as recruitment,
turnover and engagement.
This issue of data reliability is the
driving concern for many at the
aspiring level.
At the more advanced level
where data reliability is, if not
established, at least a major priority
the focus is on credibility. How
can the data they are collecting
help to inform business decisions
by demonstrating the impact of
people and performance? How
does this data fit with other key
business data streams? Is this data
reliable enough to be mined for
predictive insight?
References
ACCENTURE. (2013) How well do
you know your workforce? [online]
Accenture Outlook. No 3.
Available at:
http://www.accenture.com/us-en/
outlook/Pages/outlook-journal2013-how-well-do-you-know-yourworkforce-analytics.aspx [Accessed 1
November 2013].
ANDERSON, C. and SALLY, D. (2013)
The numbers game: why everything
you know about football is wrong.
Harlow: Penguin Books.
BERSIN BY DELOITTE. (2013) High
impact talent analytics: building a world
class HR measurement and analytics
function [online]. Oakland, CA: Bersin
by Deloitte. Available at:
http://marketing.bersin.com/highimpact-talent-analytics.html [Accessed
1 November 2013]
BERSIN, J. (2012) Big data in HR:
building a competitive talent analytics
function the four stages of maturity
[online]. Oakland, CA: Bersin by
Deloitte. Available at: http://www.bersin.
com/Practice/Detail.aspx?docid=15430
&mode=search&p=Human-Resources
[Accessed 1 November 2013].
BOUDREAU, J. and JESUTHASAN,
R. (2011) Transformative HR: how
great companies use evidence-based
change for sustainable advantage. San
Francisco: Jossey Bass.
BOUDREAU, J. and RAMSTAD, P. (2007)
Beyond HR: the new science of human
capital. Boston: Harvard Business
School Press.