Basic Steps for New Hire Turnover

1. Total up the number of people you have hired in the last two years (or whatever period you are concerned with); don’t worry about those who were hired before your cutoff. Just filter them out.
2. Get the total number of new hires that also voluntarily left your organization in that same time window.
3. Divide the number of leavers in Step 2 by the total number of new hires in Step 1 to get your new hire turnover rate.

Basic Steps for New Hire Retention

Retention is just the complement of turnover.

1. Total up the number of people you have hired in the last two years (or whatever period you are concerned with); don’t worry about those who were hired before your cutoff. Just filter them out.
2. Get the total number of new hires who have stayed with your organization in that same time window.
3. Divide the number of stayers in Step 2 by the total number of new hires in Step 1 to get your new hire rate.

It’s that simple.

What Time Period is Best?

There is no single answer because it depends on your business question. To help you get a feel for your data and your organization, I would suggest that you plot the duration of employment for those who leave voluntarily using a histogram or a boxplot.

You might find that most of the quitting is done within the first few months…or perhaps most of it occurs after 18 months…or perhaps it’s evenly distributed.

Why is this important?

Because the timing of when people quit relative to their hire date might tell you something meaningful about underlying hiring, turnover, and retention challenges.

For example, the steps you might need to consider if everyone quits in the first month would likely be different than if you find people are mostly leaving after a year or two. One pattern might signal poor hiring practices or poor pay while another might signal frustration with limited opportunities for career growth.

What Counts as a “Voluntary” Departure?

Depending on the type of organization you work with and the kind of roles you are filling, you might need to carefully consider your definition of “voluntary” quitting.

For instance, sometimes employees just stop showing up. That might be considered a form of quitting even though they might technically have been fired. In other cases, there is simply a broader pattern of not being reliable, leading to involuntary removal. You also might need to consider leaves of absence and remove those from the calculations.

Regardless of what you decide, the key is just be transparent and consistent with what you are measuring.

Finally, I would encourage you to keep a running calculation, performing it every quarter so that your new hire retention window moves along over time and you can see trends emerging.

That said, I would caution you against going back too far in time. Looking at a 3-year window that began in 2013 and ends in 2016 might have very little relevance to decision making processes in the current economic and labor climate.

Examples in R and Excel

Let’s first generate some data to work with.

set.seed(42)

emp_id <- seq(1,1000)
hire_date <- sample(seq(as.Date('1998/01/01'), as.Date('2018/01/01'), by="day"), 1000)
term <- sample(c(0,1), 1000, replace = T, prob = c(.9, .2))

emp_dur <- as.integer(1000)
for (i in 1:1000){
  if (term[i] == 1){
    emp_dur[i] <- rnorm(1, 180, 40)
  }
} 

term_date <- hire_date + emp_dur

d <- data.frame(emp_id, hire_date, term, term_date, emp_dur)

Our data is a mix of people who have left and who are still here. It also has a range of hire dates.

d2 <- subset(d, hire_date >= as.Date('2016/05/01'))

mean(d2$term)

## [1] 0.2278481

summary(d2)

##      emp_id        hire_date               term       
##  Min.   :  2.0   Min.   :2016-05-04   Min.   :0.0000  
##  1st Qu.:241.0   1st Qu.:2016-10-04   1st Qu.:0.0000  
##  Median :542.0   Median :2017-02-19   Median :0.0000  
##  Mean   :512.7   Mean   :2017-02-24   Mean   :0.2278  
##  3rd Qu.:790.0   3rd Qu.:2017-07-31   3rd Qu.:0.0000  
##  Max.   :979.0   Max.   :2017-12-14   Max.   :1.0000  
##                                                       
##    term_date             emp_dur      
##  Min.   :2016-11-10   Min.   : 90.91  
##  1st Qu.:2017-05-07   1st Qu.:155.68  
##  Median :2017-07-22   Median :174.86  
##  Mean   :2017-08-12   Mean   :179.06  
##  3rd Qu.:2017-12-01   3rd Qu.:207.33  
##  Max.   :2018-05-19   Max.   :243.86  
##  NA's   :61           NA's   :61

library(ggplot2)
library(wesanderson)

ggplot(data = d2[d2$term == 1,], aes(x = emp_dur)) + ggtitle ('Distribution of Employment Duration for Leavers') +    xlab( 'Employment Duration') + geom_histogram(fill = wes_palette("Zissou1")[2], binwidth = 25, color = 'black')

Calculating New Hire Turnover and Retention in Excel

I’m going to export the data we created to a csv and then show you how to calculate new hire turnover in Excel.

write.csv(d, './new_hire_retention_CSV.csv')

First, we’ll open up our data in Excel and create a pivot table.

Then we’ll “hire date” to the filter column and insert a timeline filter

Finally, we can calculate the average of our term column to give us the percentage of new hires that voluntarily termed in that period (23% as expected), yeilding a complementary new hire retention rate of 77%.

Summary

New hire retention and turnover are essentially HR metrics but just as importantly, be sure to look at the distribution of voluntary terminations relative relative to the hire date. Doing this will help move you from looking at a single number to talking about your hiring, your people, and the processes that drive their employment decisions.