Talent Acquisition

1. Time to Start (Average time it takes to fill a position)

Time to start is calculated by taking the total days positions are open and
dividing them by the number of positions filled.

This metric can provide insights on the effectiveness of your employment

brand, marketing efforts, and application process. "High" numbers can also
provide perspective on your interviewing process. You may have too
lengthy of an application process, too many interviews, or a weak candidate
experience.

2. Time to Productivity (Average numbers of days to satisfactory

productivity)

Time to productivity is calculated by taking the number of days between the

new employee's start date and the point at which he or she reaches
satisfactory productivity and dividing it by the number of positions filled.

This metric is a vital indicator of how well your onboarding

program is performing. Also, outcomes can be an indication of the quality of
your recruitment process and the caliber of the applicants you're attracting.

3. Turnover Rate (Rate at which employees are leaving the organization)

Turnover rate is calculated by taking the number of departures during a

specified period (a year, quarter, or month) and dividing it by the
average number of employees during the same period, and then multiplying
by 100.

When focused on a target group (e.g., high performers, Millennials, low

performers, or critical positions), this metric provides insights into the
effectiveness of your performance management, development, or culture
initiatives.
4. Cost Per Hire (Average cost incurred with a new hire)

Cost per hire is calculated by taking the cost (marketing spend, recruiter fees,
or relocation costs) and dividing it by the number of new hires.

Too high of an external cost could be an indicator that you need to invest in
internal recruiting resources such as training or technology, hire an
extra recruiter, or reevaluate the mediums you've selected to market your
positions.

5. Acceptance Rate (Percentage of candidates who receive offers

and accept)

Acceptance rates are calculated by dividing the total number of candidate

acceptances by the number of offers extended, and then multiplying by 100.

This number will let you know if you need to improve your candidate
experience, benefit packages, or the competitiveness of your offers.

Learning and Development

6. Readiness (How ready the organization is from a human capital

perspective to execute the business strategy)

Readiness is calculated by taking the total number of vacant positions divided

by the total number of approved positions and then multiplying that number by
the total number of employees with desired competency ratings divided by the
total number of employees who have received a competency assessment.
Then multiplying the answer by 100. Let me simplify: (Vacant Positions/Total
Positions) x (Employees With Desired Competency Rating/Total Assessed) x
100.

SHRM recommends you only perform this calculation for key roles. It's a vital
metric for determining whether your workforce is aligned with your business
trajectory.
7. Training Participation Rate (Percentage of employees participating in
development opportunities)

Training participation rate is calculated by dividing the number of employees

who participated in a training opportunity by the number of employees who
were eligible, and then multiplying by 100.

Results will help you determine if you're offering the "right" type of training, if
you're using the best delivery medium, or if you've effectively communicated
the opportunity.

Performance Management

8. Average Performance Rating (The average performance rating for a

selected group of employees)

The average performance rating is calculated by dividing the sum total of all
performance ratings by the number of employees who received a rating, and
then multiplying by 100. I would suggest taking this number and comparing it
against a larger distribution.

Tracking this metric will let you know if your performance management
programs are working effectively and if any teams are falling behind the
curve.

Benefits

9. Benefit Participation Rate (Percentage of employees participating in a

particular benefit plan or program)

Benefit participation rates are calculated by dividing the number of employees

enrolled in a plan or benefit by the number of employees eligible, and then
multiplying by 100.
Every employer benefit comes at a cost. If employees aren't using some of
them, then cut ties and move on. Reinvest the money into plans that are more
widely used.

Organization Effectiveness

10. Employee Engagement (Degree to which employees are engaged with

and committed to the strategy and objectives of their organization)

Calculating engagement is tricky. Most organizations invest in a third-party

resource that typically includes an extensive survey -- it's worth it. A high
level of engagement is a critical trait found in high-performing organizations.

11. Workers' Compensation Claims (Measure of workplace safety)

Analyzing your workers' compensation claims (amount and cause) will ensure
you provide and maintain a safe working environment for your employees.

Volume 0%
01:19
01:19
Step 1. Define The Goal
Our hypothetical company wants to apply data science principles and steps
to deal with a critical HR issue: employee churn. It realizes that when good
people leave, it costs far more to replace them than providing them with
some incentives to keep them. So it would like to be data-driven in the HR
decisions it makes with respect to employee retention.

1. What proportion of our staff is leaving?

2. Where is it occurring?
3. How does Age and Length of Service affect termination?
4. What, if anything, else contributes to it?
5. Can we predict future terminations?
6. If so, how well can we predict?

Step 2 – Collect and Manage the Data

Often the data to analyze the problem starts with what is currently
readily available. After some initial prototyping of predictive models,
ideas surface for additional data collection to further refine the model.
Since this is first stab at this, the organization uses only what is
readily available.

After consulting with their HRIS staff, they found that they have
access to the following information:

 EmployeeID
 Name
 Birth Date
 Original Hire Date
 Termination Date (if terminated)
 Age
 Length of Service
 City
 Department
 Job title
 Store Name
 Gender
 termination reason
 termination type (voluntary or involuntary)
 Status Year – year of data
 Status – ACTIVE or TERMINATED during Status year
 Business Unit – either Stores or Head Office
First Look at The Data – The Structure
Second Look at The Data – Data Quality
Third Look at the Data – Generally What Is
The Data Telling Us?

Step 3 – Build The Model

Attrition % Attrition % Attrition %

FY(2015-16) FY(2016-17) FY(2017-18)

Tenure wise

Age wise

Reason wise

Evaluate Models
Can we predict?
In a word ‘yes’.

How Well can we predict?

In two words ‘fairly well’.

Deploy Model