To remain competitive, we need to use metrics-driven data to build credibility as a strategic partner. photo | fotosearch
Companies rarely make their strategic direction public, at least
the finer grains of it. It is however possible to introspect the
strategic direction that a company is taking by looking at the job they
are advertising. One such noticeable trend is the increase in Job
adverts for data science talent . One particular job listing that caught
my attention is a people data scientist listed by Pay Pal, below is an
excerpt from their website.
“As the People Data
Scientist - Principal, you will work with the and team lead and other
Data Analysts, Data Engineers and Data Scientists responsible for
strategic analytics initiatives, research and experimentation. You will
solve essential business challenges such as: Improving collaboration
across the organisation; Predicting which employees are at risk of
leaving the organisation before they leave; Predicting which of our
1,000,000 + applicant each year would make great hires and many other;
Through data allowing employees to plan and navigate their future
careers at PayPal; Identifying what characteristics make a great sales
person, engineer, among others; and, many other key questions we will
leverage data to solve.
"Reporting to the Head of
Global People Analytics, you will be directly engage with PayPal People
Strategy, Analytics and Technology Team to solve scope complex human and
data questions, analyse, model, test, and build solutions that provide
insights and recommendations”
HR has a history of using
their gut feeling to make decisions. To remain competitive, we need to
use metrics-driven data to build credibility as a strategic partner.
Big
data for HR encompasses these three things at its core 1) predictive
analysis, 2) infusing data from other systems related to finance
customer service marketing and other functions, and having the ability
to make daily decisions based on evidence that are directly related to
HR processes, 3) Using both to then design workforce structure and
planning.
Data will talk to you if you’re willing to
listen, however, the challenge comes when data transforms into bundles
and stacks of unorganised and unstructured data sets. HR metrics and
workforce analytics are not a guarantee return on investment. You need
to be able to decipher what data is appropriate to collect. Learn how to
use this information to increase managerial decision-making efficiency.
HR business partners need many skills to perform
their job effectively. An important skill that is lacking in some HR
business partners is the ability to analyse and interpret data. With
this skill in their toolkit, the HR business partner can apply business
analytics to analyse returns on investment on an HR programme.
Analytics can be used to support decisions with data through
business cases, such as a developing a leadership programme or piloting
an employee satisfaction survey. Business analytics provide a wealth of
value that can save organizations money. By connecting HR metrics to
business performance, executives can see how HR impacts the bottom line.
There are very few prerequisites to begin predictive
analytics: a) access to workforce data; b) having a workforce issue
where understanding the probability of the future outcome is desirable;
c) access to, or the ability to resource, advanced statistical analysis
talent.
It is imperative, however, that HR teams
strive to progress in all areas of workforce analytics, rather than
focusing on just one or two areas. Proficiency in one area, like metrics
reporting, will not automatically lead to progress in predictive
modelling.
Predictive modelling is not about having
the most data at your fingertips— despite the buzz about “Big Data.”
It’s about testing the right hypotheses.
For example,
if you’re testing for attrition risks for call centre employees, you
will want to identify a set of potential reasons (or hypotheses) why
people are leaving— long work hours, low pay, difficult commutes, poor
job fit, and the like. You will need to collect the data to test those
hypotheses, and that does not mean “all possible data.”
No comments :
Post a Comment