Analysing Turnover And Retention

Why Analyzing Turnover is Not Enough December 18, 2007

“Why Analyzing Turnover is Not Enough” is a very salient topic and certainly something at the top of mind for a lot of people. We are lucky today to have with us Kelly Northrop as our speaker. She knows a lot about this topic, as she looks at this data constantly. Kelly is an analytic consultant for the talent management division of Kronos. She works with Kronos’ clients and internal selection services team to develop and apply innovative selection and measurement techniques that drive workforce performance.

KN: We would like to take you through some discussion about turnover and retention as measures that you can use to look at the rate of departures and separations within your company. Some of the discussion topics that we have today and just a preview of what we will be going through in the slides that we have include first, what is turnover? We will give a basic overview of turnover, how you calculate it, what some of the uses of it are. Second, we will talk about what is “wrong” with turnover? Although turnover is a very broadly-used global metric, there are some shortcomings to it, and we will talk about that. Third, we will talk about new hire retention. This is sort of an alternative metric that you can use as an alternative to turnover. We will talk about what it is and

why you might want to use it as well as some of the caveats for using retention instead of turnover. And lastly, we will talk about the economic impact of a separation.

POLL QUESTION #1 1. How often do you measure turnover or retention for your organization as a whole? a. b. c. d.

Monthly Quarterly Annually Do not measure

Poll Response: It looks like most of you, about 40%, measure turnover on a monthly basis which is quite good and then we have another 30% who measure quarterly and then we have sort of a breakdown of the rest between annually and those of you who are currently not measuring turnover. Based on our experience at Kronos with primarily hourly workforces, I would say this is a pretty typical result of what we usually see in working with our clients.

What is Turnover?

When you think about turnover, it’s pretty simple, right? It’s a simple calculation, separations divided by headcount multiplied by one hundred and you have a turnover percentage, very straightforward to calculate.

However, when you start to think about it, it is actually kind of complicated. There are a number of factors that play into turnover and I’ll go over some of these here. Let’s start with headcount on the lower left. Headcount is really a measure of the operational capacity that is required to support an organization. There are many breakdowns that you can look at in terms of headcount, different job families, employees that work fulltime versus part-time. Headcount can swing widely during different seasons, say for example holiday season, summer season, and hiring season, all of these things can affect headcount at any one time. As well, within sort of the multiunit industries and retail industries, we have store openings. If a company is growing quite a bit and adding locations that can really affect headcount by increasing it rapidly over a short period of time. On the separation side, we have also got both voluntary and involuntary separations. These are usually categories that attract within an HRIS or payroll system. Under voluntary and involuntary, we have categories that you might consider good terminations or bad terminations. So, there might be a good termination, which in the context of this webinar we are using a good termination as a termination that is to the benefit of the company. A bad termination is one that is not good for the company. In

other words, the person you would have chosen not to lose from the company if you have the option. So, within both voluntary and involuntary, you can see there are a number of different separation reasons that we have. Under involuntary especially we have got what you would typically think of is firings for misconduct or negligence. We have also got structural reasons for separations that would include lay off, downsizings, restructurings, and that type of things. Next, we have cost. So, when we look at turnover, there are different types of costs that you might want to calculate. Some are easy to calculate and some are difficult. The replacement cost is the pure hard cost of losing somebody. What a cost is to rehire and retrain a replacement. There is also the opportunity cost, which is significantly more difficult to calculate, but which we think is definitely worthwhile calculating in that opportunity cost is what happens when someone leaves the organization. There is the different productivity and other people basically have to make up for the lack of productivity in that head count, and we will talk more about that later. Another element of turnover is time. So, again our polling question, do you look at it monthly, do you look at it annually, other financial periods that are of particular interest to your company, sort of fiscal year for example, and then lastly we have slices. Slices would be field versus corporate employees trying to figure out if there is a correlation or causation relationship between different things say, for example, profit that is the EBITDA that you see there on the third bullet. Is there a relationship between profit and turnover? Are more profitable parts of your business associated with higher or lower turnover than others? And then especially in the field contacts, we have manager tenure, so that is also a question of correlation versus causation. What is the tenure of the manager and do hired tenured managers tend to have lower turnover and vice versa? So again, these are some of the things that play into turnover and they most definitely complicate the picture from the standpoint that some of them can serve to increase turnover and some of them will serve to decrease turnover.

“What does it all mean?” Slicing

Trending

Ar ea

P er ce nt o f P op ul at i on

Co r p HQ

3. 2%

10. 02%

0. 32 %

Co r p Re m ot e

. 6%

11. 69%

0. 07 %

Di st r i bu ti on

9. 9%

24. 67%

2. 44 %

Fr an chi se F i el d

. 4%

20. 00%

0. 08 %

M ar ket i ng F ie ld

. 1%

7. 14%

0. 01 %

St or e O f f ic e

. 6%

200 3T ur n ove r % F or

25. 00%

I m pa ct %

0. 15 %

St or e s AM

8. 0%

79. 4%

6. 35 %

St or e s GM

4. 0%

48. 02%

1. 92 %

St or e s Dr i ver

50 .6 %

104 .3%

52 .7 8%

St or e s CS R

22 .7 %

149 .04 %

33 .8 3%

Gr o up

2 00 .0 1 80 .0

1 74 .2

1 65 .5

1 60 .0 1 40 .0

1 59 .2

1 53 .6

1 28 .1 1 1 8.0

1 20 .0 1 00 .0 80 .0

99 .1

60 .0 40 .0

1 0 0.3

1 19 .5 78 .2

5 8 .6 46 .2

4 2.9

11 1.0 79 .4 48 .0

20 .0 0 .0 20 0 0

20 01

2 00 2

20 0 3

“What causes change?”

“What does it cost?” What is

“Do we have a problem?”

“Causation or correlation?”

“How do I reduce it?”

em y HRIS syst Why can’t m ed! ne I at wh e give m

Which leads us to the question of what does it all mean? We talked a little bit about trending, we talked about slicing, how do I know if something is just correlated if I look at two things and I correlate something with turnover? How do I know if those things are just happily moving together or whether one of them is actually causing the low turnover

or the high turnover that you might be seeing? What does it cost? So again, we talked about the hard cost and the soft cost of turnover and then also, do we have a problem? Is it good or bad turnover that you are seeing when you calculate turnover, which positions are driving it, and we will talk a lot more about which positions might be driving turnover and how you analyze turnover and retention by different position categories, which kind of leads to the basic question of what is the root cause of turnover. Once you are able to dig into this metric a little more, you dig into retention, which we will see later. It will help you know basically where to apply resources and fixating problems and that’s really what it’s all about, so although it’s good to know the overall metrics, what you are really looking for is data that you can act upon.

What is “Wrong” with Turnover?

There are some underlying problems. First would be headcount. What we see here is separations divided by headcount is the basic way the turnover is calculated. Number of separations divided by headcount during that period of time. On the headcount side, turnover treats all headcount the same. It doesn’t matter whether it’s a seasonal hire or seasonal employee. It doesn’t matter whether it’s a vice president or a clerk. Everybody is treated the same in a turnover calculation. It also hides the critical fact that different job families have different base rates of retention. Different job families and different employee types have different rates of turnover and you would expect that to happen and you would also want that to happen depending on the dynamics of your business. On the separation side, turnover also treats all separations the same and what you kind of know empirically is that the impact of separation varies by position. Obviously the impact of a vice president leaving has a greater effect on your organization than the impact of an entry-level employee. It also assumes that all separations are the same in terms of a good or bad termination. Turnover treats all of this the same and then also as we will see in a bit reasons for separation vary by tenure. There are different reasons that tend to come into separations early in an employee lifecycle versus later in the employee lifecycle. The question with all of this is what you are actually measuring when you blend all these together. What is the real effectiveness of this metric? In effect, you are combining all of these effects that may or may not be working against each other into one overall metric and it’s very difficult to take action on that.

Turnover = Volatility Annual View

Daily View 57,000

100,000

Headcount

75,000

56,500

Headc ount 50,000

56,000 25,000

Jul

Aug Sep Oc t Nov Dec

Fr i Sa t

Feb Mar Apr May Jun

Fr i Sa t Su n M on Tu e W ed Th u

Jan

Su n M on Tu e W ed Th u

55,500

0

• Turnover determines the “spikiness” of short term changes in headcount • Greater spikiness means a larger reserve of labor is required to maintain the operation • Operational impact: managers have to smooth out the spikiness – – –

Overtime costs Scheduling revisions Training costs

Turnover is really a measure of volatility. What you are seeing on this slide, on the left you are seeing annual view of headcount, and on an annual basis it looks pretty smooth and consistent across the months, but when we look at the daily view, which is what you see on the right, you see that on a daily basis, headcount actually fluctuates quite a bit from day to day. It’s actually the turnover that is occurring that makes that headcount fluctuate and it determines the “spikiness” that we see in the graph that occurs from day to day. Effectively, what that means from an operational standpoint is that the greater the spikiness is and the more the headcount fluctuates from day to day, the more staff that you need to buffer those changes and the turnover that is occurring. For managers particularly in an hourly employee contacts, but again this is true as well with corporate or salaried jobs, managers are the ones that have to deal with the spikiness. So, if you are looking at hourly employees that means they have overtime cost of departures that they didn’t anticipate. They have scheduling revisions, training costs. These all play into managers dealing with the turnover that occurs in their environment from day to day, and

most of the time in hourly the turnover is unanticipated. We will look a little bit at reasons why turnover occurs at different stages in the employee lifecycle, but basically it’s the managers in the field that have to deal with the turnover that is occurring, and the way they do that is by invoking things that are costly to the organization and that would be the overtime in the training. It is also important to note that the daily changes are driven by turnover, so the daily fluctuations and the spikiness and the headcount is driven by the turnover, but long-term changes and headcount are usually structural. When opening a new office or in manufacturing opening a new plant, think about major structural changes, technology, outsourcing, whatever it might be that changes headcount in the long term, those are usually structural changes, but it’s not really the turnover that changes the headcount on a structural basis.

Next, we have an illustration of what we mean when we talk about volatility and one of the ways of calculating cost of turnover is an HR transaction. This is again a pretty easily calculated cost, it can vary in the hourly environment. We talk about it anywhere from probably $1,000 to maybe $2,500 in salary depending on if you are paying recruiting fees or anything like that. It can mushroom into much, much larger dollars, but basically there is a cost of the separation and that comes in the form of an HR transaction that has to be made. What we see here is an example and again this is a typical retail example where we have two locations, we have got location A on the left and location B on the right. Both of these locations, significantly they both have a headcount of 10 people. For simplicity sake, we assume that all 10 of those people are hired on January 1st, and what we see in both of the boxes below location A and location B is the number of employees that cycle through that location in the course of a year. So, on location A, what we see is that we start off with 10 employees and halfway through the year we have to replace all of them, and what this means is that we have separations of 10 on a headcount of 10 and so the turnover is 100%. On location B, employees actually turnover at twice the rate and what this means is that there were 10 initial employees that have to be replaced twice during the course of the year. So, every 122 days. For location B, we had 20 separations on a headcount of 10 and that’s how we get the 200% turnover. You can see just visually the number of employees that are cycling through location B is much greater than location A, and what this means mathematically is that location B has 30 employees that cycle through that location

during the course of the year. We call that the absolute employee count and location A has 20 employees that cycle through in the course of the year. So, that difference of 10 HR transactions between location B and location A really adds up over time and when you think about the impact of that across many locations and many employees over time, it can really add up just purely in the cost of HR transactions. Obviously, that is not the only cost you want to consider, so later on we will talk a little more about the overall economic impact of a separation.

What we would like to look at now is separations by tenure. This is actual data from Kronos to clients. This is a base of our retail customers and this is the number of separations plotted against the days of tenure at the time of separation. So, you will see across the bottom of the graph, it says “days of tenure at time of separation” and then up the Y-axis is the “number of separations” that occurred. What we look for here is the slope of this look fine, and if it is very steep, it means that a lot of people are separating early on in their tenure. There are kind of what we call early quit in the hourly environment, and you can see this is a plot of retail non-exempt employees. We did exclude seasonal employees because that would change the picture quite a bit. Day 13 is the most common “quit day,” so really what that means is there is a lot of people that come on the job, decide it’s not for them and they quit within a couple of weeks of starting. Similarly, we can see that 75% of all separations occur within the first year, so that really gives you a feel for the amount of time that an organization with hourly employees has to really effect that employee’s experience within the company. It suggests that onboarding and early training are very important and in fact at Kronos we have done some research with some of our clients to shed some light on that. What we have found is that in the hourly environment employees although sometimes they appear to be not caring, they actually care quite a bit about being trained adequately and they are actually very uncomfortable with being sort of thrown into the mix without a lot of training. They really want to be able to do their jobs well and it’s very frustrating to them when they feel that they have been undertrained. So, this is kind of some interesting primary research that we have done with some of our clients at Kronos. Similarly, we see here 33% of all separations occur within 90 days. So again, not a lot of time to get the mindshare of those employees once they join the organization. And

again, as it says in the box on the lower left there that turnover can occur for different reasons at different stages of the lifecycle and it also suggests that you can target different interventions or different programs at different stages of the employee lifecycle.

POLL QUESTION #2 2. Does your company differentiate between “good” turnover and “bad” turnover? a. Yes, we track which separations are good for the company and which are bad for the company b. Yes, but only for certain positions / roles c. No, we do not distinguish between types of turnover

What causes attrition in hourly jobs? •

Term code analysis of payroll data

•

Within the Poor Fit Category:

•

Early quits are predominantly Job Abandonment (43% versus 17% in later stage terms) Dissatisfaction with hours becomes a much more significant factor over time (78% of quits in later stage)

•

Termination Classsification Misconduct Negligence No Data No Fault Personal Poor Fit Professional Grand Total

Total 44 22 127 3 18 427 266 907

16 14 12 10

Count 8 of Terms 6 4 2 0 1

13

25

37

49

61

73

85

97

109 121 133 145 157 169 181 193 205 217 229 241 253 265 277 289

Days of Service

What we are looking at is the question of what causes attrition in hourly jobs. This is a calculation of aggregate data from our Kronos clients. Again, these are primarily hourly employees. What we are looking at is reasons why people leave the organization. This is term code information that is populated in payroll data systems by managers in the field and provided that our clients and possibly your company comply with filling this information in it will be available to us to analyze post departures, so we can look at why people are leaving. There are other ways of getting this information, for example, conducting a survey of employees who have exited, that type of thing. This is a fairly quick and easy way of looking at it. Sometimes this data is not filled in by managers and in those cases they can be a little more difficult to determine why people are leaving in the organization, but we can look at terminations in the poor fit category. What we find is

that in this hourly context, 43% are in the early stage terminations of primarily job abandonment. If you recall, we were looking at separations by tenure and we saw that the more common day of departure for hourly employees is day 13, which means that people are basically walking off the job after a couple of weeks. Similarly, this is kind of mathematically what that means, so there are 43% of the departures in that early stage, that first box there on the graph of job abandonment whereas in later stage terms that was only 17%. Similarly dissatisfaction with hours becomes a much more significant factor over time, and we found that 78% of quits in the later stage were due to dissatisfaction with hours. Now in the hourly job or hourly workforce, getting the schedule that is requested, especially for more senior employees, they expect to be on the best schedules as they gained seniority within the organization. Getting the schedule that they requested is absolutely key to their satisfaction on the job. So, it’s no surprise. Their dissatisfaction with hours is a significant factor in those quits during the later length of tenure. This same analysis could be done for corporate or salaried jobs, obviously job abandonment is less of an issue in the salaried contacts, but there is the issue of that poor fit hire. The person that may leave within a month or two months or three months of their initial hire dates, or up to, for example, the 90-day probation period. So again, the reason we look at this is to know why people leave at different stages of the lifecycle. Once we know that, that helps to target programs appropriately to different types of employees.

In the aggregate, labor market conditions drive turnover B u reau of L ab o r S tatis tics - D iffu s ion In d ex H i storic al In d ex & KT M D F orecast (12 M on t hs ' E n d in g ) In d us try = T otal P rivate

1 00 90 80

Job Creation = Turnover increasing

70 60 50 40 30

Job Destruction = Turnover decreasing

20

12 M o nth D i ffus i on

2 0 08 S e p

2 0 0 8 M ay

2 0 0 8 J an

2007 Sep

2 00 7 M a y

20 0 7 J a n

2 0 0 6 Se p

2006 May

2 00 6 J a n

2 00 5 S e p

2 0 0 5 Ma y

2005 Jan

2 0 04 S e p

20 0 4 M a y

2 0 0 4 J an

2003 Sep

2003 May

20 0 3 J a n

2 0 0 2 Se p

2002 May

2 0 02 J a n

2 00 1 S e p

2 0 0 1 M ay

2001 Jan

2000 Sep

20 0 0 M a y

2000 Jan

1999 Sep

1999 May

1 99 9 J a n

1 9 9 8 S ep

1 9 9 8 Ma y

0

1998 Jan

10

F o r ecas t

Source: Bureau of Labor Statistics – Diffusion Index. Historical index & KTMD forecast (12 Months’ Ending). Industry = Total Private.

One of the other big factors that we know is out there, but sometimes getting data on it can be a little tricky, is the labor market. What we find consistently at Kronos is that it is the labor market that drives turnover in the aggregate. Now obviously any one company’s situation can vary drastically, but when we look across all of our clients and we look at the economy, we really see a significant correlation between those two things. What that means is that clients of Kronos, or at least the ones who have been able to aggregate the information, are definitely subject to the wider labor market. What we see here on this graph is the Bureau of Labor Statistics – Diffusion Index. It measures whether jobs are being created in the economy or being taken out of the economy and that red line at 50 is where the index crosses from job creation territory into job destruction territory. This is plotted out for the last approximately 10 years and you can see the last big recession we have starting in 2001, there was quite a bit of job

destruction. In other words, net negative job creation within the economy means that people have few options looking for their next job and they are much less likely to switch jobs. So, what we see in that case is a decrease in turnover. On the flip slide when the diffusion index is above the 50, we see job creation and that means that turnover is going to increase all. People have more options in the job market and they are more likely to change jobs and so we see turnover increasing. What you see on the very right there in green is a forecast from the Kronos analytics group based on history where the diffusion index might end up in the next year. So, still in positive job creation territory but only to a slight degree and certainly less aggressive job creation than we have seen over the last couple of years.

What is New Hire Retention and Why do We Use it?

I have been talking about this metric retention and how possibly it’s better to use than turnover for all the reasons that we talked about, why turnover can be complicated when you start to get under the surface of this. That easy initial calculation.

What is new hire retention? •

It targets one specific element of “Separations” – the outcome of new hire decisions

•

It is a critical measure of hiring success – allowing for empirical measurement of new hire programs • • •

Selection Onboarding Training

•

It has direct productivity implications when tied to “time-to-competency” and hiring break-even

•

It is easier to measure than turnover

First, we will talk about, what is new hire retention? One of the reasons we use this metric of new hire retention at Kronos is because it targets a specific element of separations and that is the outcome of a new hire decision. In that respect, it definitely targets the hiring success that you are seeing within your company. It allows for measurement of the new hire program. In other words, you can put a new hire program in place and then measure retention afterwards to see if retention went up or down. Because it is focussed on new hires, it allows us just to isolate the effects we are seeing to that group of employees. When we look at turnover, we know that separations are occurring but we don’t distinguish between did someone leave who was on the job for 30 days or did someone leave who was on the job for three years. Those are all lumped together. With retention, it’s specifically looking at what happens to people after they are hired. From that standpoint, retention can help shed light on the selection process, so it’s sort of your interviewing, selection assessment program that you have sustained, your onboarding process as well as training, and I talked a little bit earlier about the importance of training that is complete and successful early on in the employee’s tenure. New hire retention also has productivity implications when we tied it to "time-tocompetency" and break-even and we will talk more about that coming up, and then lastly it is easier to measure than turnover, which we like. Turnover has a few elements to it. You have got the separations element and the headcount. Those are basically two data streams that you have to manage and measure and calculate. With retention, all you need to know is when someone is hired and then when they left the company.

As we see on this next picture of new hire retention results, retention literally is the percentage of a group of new hires who are still on the payroll after a specified period of time. What we do is take a group of new hires and then at different points in the future basically after they were hired, we look at what percent of that group is still on the payroll. That’s what you see across the bottom where it says “Days Post Hire” and it has 30, 60, 90. Those are one-month increments, that is typically what we use for hourly jobs. You could do the same analysis for corporate or salaried positions, you would obviously use different time metrics, so for a corporate or salaried job, you might start with 90 days for a probationary period and then measure it to look at retention every three to six months or annually after that, depending on what types of positions you are talking about. The line on this chart is plotting the percent of that group of new hires who

are still on the payroll after those time periods. So, what you see is that after 30 days, 88.3% of those new hires were still with the company and then similarly after 330 days out at the end there that there were only 38.4% of that group of new hires still with the company. For that reason, this curve can never go up, it can only go down because you are looking at a fixed set of new hires and then plotting how many of them are still with the company. So, a true attrition. Naturally, there will be a few of them left on the payroll as you march forward in time. The median retention is 203 days, so an easy way to think of this is that any new hire basically has a 50-50 chance of making it to 203 days. It’s kind of an easy way of thinking about that. This is a typical representation of new hire retention results. This is something that we use frequently in our analytics group and there are a number of different ways that you want to work at retention and that’s what we will talk about next.

Just as you can’t lump all decisions together for turnover analysis, you wouldn’t want to lump them altogether for retention analysis either. What we have here is a variation on the graph that you just saw. Each one of these lines is a different position within a company and this is an example of Kronos clients. You can see there are positions up at the top with that blue line that have very high retention and the positions with lines down near the bottom have lower retention. As I pointed out before, the lines are always going down from the upper left to the lower right as attrition occurs in these different groups. There is really quite a big difference between group A and group B. Group A at the top probably corresponds to store manager or store director or department manager, something like that in the retail context. Group B is probably an entry-level clerk and you can see there is a huge difference in their retention. After sic months, 85% of the employees in group A were still with the company whereas in group B only 33% were still with the company. So, naturally just as the turnover with retention, you would not want to combine these two groups together into one metric.

Another way that we like to look at retention is by geography. So again, what we see on this graph are different locations of a multiunit organization where everyone in the blue vertical lines represents one location and they are plotted from low to high in terms of their retention. What we are looking at here is 90-day retention and the median site retention at 90 days is 54%. So that means that half of the sites had retention higher than 54% and half of them, their retention was lower than 54%, but what you can see is a huge variation between locations. There are some down in the thirties, 30% retained at 90 days, which is quite low and then there are some close to 100% retained in 90 days. What this type of information helps drive is knowing where to focus resources and where to focus your efforts. You could do the same analysis for similar departments or offices in other companies; this does not apply only to multiunit retail businesses, but looking at this geographic basis or this unit basis is very helpful in knowing where your issues and where your opportunities are. As well, we found that especially with risk factor retention that there is some opportunity for the lower performing sites to learn from the higher performing sites. Often there are best practices they are taking place at the higher performing sites that otherwise might not have been teased out and made available to educate the rest of the organization, if those sites weren't identified as being the high performers initially. Some of the reasons for low retention or high retention for these variations that we found can include the quality of the management at each of these locations. It can include the experience of the management, so sometimes it would be that managers with a lot of tenure would have high retention amongst their staff at that location. We talked about the labor market, so absolutely the local labor market plays into retention at that location. We find when we work with some clients that low retention or high retention is just a function of where that location is located and it can be difficult to change that. Another thing that can play into it is compensation level, so depending on the market that these different locations are in, there might be different demands on compensation and sometimes there can be some disparity between what a company is paying and what the local going rate is. These are some reasons why some locations would be at the bottom and some at the top.

What’s the Economic Impact?

What we will be showing you in this section is a conceptual way to think about the cost of the separation to your organization.

POLL QUESTION #3 3. Have you quantified the cost to your organization of an employee separation? a. Yes, including both hard and soft costs b. Yes, hard costs only c. Have not quantified the cost

Poll Response: It looks like the majority has not quantified the cost to their organization of an employee separation. I think you are in good company, that is fairly typical as well as for the hard costs we see 22% for the hard cost to be known, that is also typical. So again, depending on your environment, it can be easier or more difficult to calculate these things and it’s really on an individual company basis whether that is something that is practical to calculate, especially when we are talking about the soft cost.

Let’s move on to a way of thinking about this issue. This is what we called a productivity impact of a separation. A couple of concepts that we have here are fully effective employee status and that’s what you see the acronym there as FEE. That represents the level of productivity that is achieved by your basic employee who is up to speed. So, it doesn’t reflect the top performers, but it does reflect average expected performance after someone who has had the opportunity to be trained and come up to speed on their job, so that is FEE. What we see going from left to right across time is we have a fully effective employee who terminated and that’s what this term as a red star there. Their productivity or the productivity of that headcount basically goes to 0 because we blocked that percent. Productivity stays at 0 for a while, while we have a time to hire, that is basically the difference between when the former employee leaves and the replacement is hired, and I think we are probably all quite familiar with time-to-hire as a metric. That can be increased any number of ways, technology is one of them, especially in hiring practices, that type of thing or other ways to decrease time-to-hire and then we have time-to-competency. So, that is represented by the number 2 there. This is basically the learning curve, so what we have here is the new hire has made that green star on the bottom and then they take certain amount of time for that new hire to come up to speed. Depending on the type of position that you are talking about, this can be easier or more difficult to calculate. In the analytics group at Kronos, we have done some calculations for clients and found that for example a typical front line or retail front line salesperson, in a condition where it might take about 120 days to come up to speed, that entirely depends on a position you are talking about and the environment of your company. Obviously for higher-level salaried jobs, this time to fully effective status will take much, much longer, so it really depends on the jobs. What we see from a practical standpoint is that the departures are constant challenges for managers, what they are trying to do is compensate for departures of the experienced employees, as the new employees are brought up to speed. If you think about the implications of this diagram across many thousands or tens of thousands of employees, if you have high turnover or low retention, so new hires don’t stay very long or you have a high separation rate, it means that there will be a large number of employees on the payroll who are not up to fully effective status. When you aggregate that across the entire and full base that

means that a larger percentage of the overall employees are not at fully effective status and basically the average productivity of the organization is decreased, so this is kind of a way to think in both the specific instance and in the aggregate of what that means when somebody separates from the organization.

How does hiring play into this? $

Cumulative Employee Contribution

Break-even

X

$0

Up to Speed

Cost of Hire

Training

Day s 30

• •

60

90

120

Each hire is an economic transaction What is your return on investment for each hire?

We talked about HR transactions before, but in the more general sense, you can think of every hire as having a return on investment and again this applies to any environment that you might be talking about, whether hourly employees or salaried. There is a hard cost of hiring as well as training and that’s what we see in this diagram on the left, so the green line where it goes below 0 is the investment that is required to bring in new employees onboard and train them up to effective status. In this example, the employee comes up to speed in 60 days. This is probably an entry-level cashier or something like that in this example. The employee comes up to speed at 60 days, but the organization does not break-even if you will on that investment made in that hire for some period after that. So, whether this can be applied practically is when an employee separates. If you have this information for a job category or a significant job category in your organization, you will know basically how much time you have to get them trained and then how long may need to be retained afterwards to pay back the initial investment in that hire. And at some point in the future, there is that break-even standpoint and obviously going into the future the longer you can retain somebody the greater the investment that was initially made.

Summary of turnover and retention •

Turnover is a critical concept that describes the relative expense of supporting operational capacity. –

As a metric it has profound limitations in driving targeted interventions.

•

Retention is a metric that is easier to calculate and can provide more straightforward information.

•

Whether looking at turnover or retention, one needs to consider position dynamics, timing, and separation reasons.

•

Considering critical milestones such as time-to-competency and completion of training provide a clearer picture of hiring effectiveness.

We talked about some of the shortcomings of turnover, especially how it is often used as a global metric across all positions, locations, and situations. We like to think if it is a concept, so is it used for concept when it comes to describing the expense of supporting operational capacity. Think about the volatility graph that we looked at before, the day to day fluctuation and headcount. It’s the turnover that causes that spikiness and it’s the turnover that allows us to know how expensive it is to maintain that headcount. As a metric, it’s difficult because of so many moving parts. As a metric, it’s difficult to know what I would do with this information. That’s why we recommend more specific and easier to calculate metrics of new hire retention. It is easier to calculate as I mentioned, you just will need that one data stream and all you need to know is when somebody joined the organization and when they left the organization. Nevertheless, whether you are looking at retention or turnover, you do need to keep in mind that different positions have different expected levels of retention and turnover, different separation reasons, different timing in terms of why people leave at different stages and their tenure and that also plays into what you would decide to do with that information. And then lastly, the feature we combine is with the things we look at in terms of the productivity impact of a separation and what is that return on investment of hire. It can really give you a clear picture of your hiring process - the costs that are involved as well as the cost of onboarding and training.