Opinion - Measuring Attrition


Staff Attrition should be one of the easiest metrics to create and trend.  It is quite simply the number of Leavers divided by the Staff Population for the period.  I have a preference for using the Staff Population as at the beginning of the reporting period, as the focus of the metric is to identify how the business has changed from that point in time.
I also have a preference for using a straight head count for both the Leavers and the Population. Again the rationale is quite simple, the issues surrounding staff attrition are the same for all staff whether Full Time, Part Time, Casual or even Fixed Term.  There is just as much effort and cost when it comes to advertising, recruiting and training regardless of contract. Advertisement placement doesn’t cost half as much for a Part Time role, nor is there half the required Training needs.  A Part Timers reasons for leaving aren’t half as important either, so why only count them as a half?

The other contentious issue is that surrounding Voluntary and Involuntary attrition. There are good reasons for and against using Voluntary Leavers as the divisor.  Again, I have a preference for using all Leavers as only using the Voluntary Leavers can be misleading.  Consider a round of Redundancies, often when redundancies are on offer, a number of staff will volunteer for redundancy as they were already considering leaving. Hence only reporting the voluntary Leavers through a period of business change could easily show an unrealistically low attrition rate. Including redundancies and showing the spike in attrition is not a bad thing as it identifies a business impact event.

The use of Voluntary Leavers only, does, however give a smoothed result that eliminates, to a certain degree, spikes due to significant business events.  The key to measuring voluntary attrition is to omit all involuntary leavers from the entire sample including the initial population.  Whilst omitting the involuntary leavers will reduce the staff population and consequently increase the percentage turnover, it will be more realistic result.

Contract Status Average Service Leavers Service Attrition%
Fulltime 7.94 5.78 22.92%
Parttime 5.36 3.16 31.82%
Total 7.83 5.68 23.22%
 
Regardless of the manner in which the Attrition Rate is determined, a more important measure is the average service length or the Leavers.  This delivers a much clearer indicator of staff retention.  It is also easy to separate the results for Voluntary and Involuntary, although it doesn’t counter the skewing of either result by staff electing to take voluntary redundancy.

A straight Attrition Metric may provide a valuable business performance indicator, but in isolation it does little in identifying areas for improvement. The real value is Staff Attrition analysis is the application of the results against a number of logical population groupings.

In the example shown, interrogating Staff Attrition by Age Band brings to light a potential area of concern.  Employees under the age of 30 have a 50% chance of leaving and will have delivered just over 2 years of service on average.
Age Band  Average Service Leavers Service Attrition%
Under 18 0.56   0.00%
18 - 20 0.67 0.26 50.00%
21- 25 1.64 2.24 43.75%
26 - 30 2.51 2.02 52.63%
31- 35 4.30 4.57 35.71%
36 - 40 5.55 3.28 21.25%
41- 45 6.24 7.62 17.86%
 
Salary Band Average Service Leavers Service Attrition%
25 - 30000 2.02 1.94 100.00%
30 - 35000 3.85 2.94 41.46%
35 - 40000 7.02 7.25 18.42%
40 - 45000 11.08 9.09 23.94%
45 - 50000 8.45 1.50 10.53%
50 - 55000 7.85 4.73 8.47%
55 - 60000 13.11 11.09 15.38%
Similarly, by interrogating Staff Attrition by Salary Band also shows that Staff paid under $35,000 have an over 40% chance of leaving and whilst the average service delivered is close to 3 years, the overall service provided by all staff being paid under $35,000 is under 4 years.

Whilst it is probably no surprise that the highest rate of attrition is with lowly paid younger staff, it is clear to see that this is by far the most problematic area.

A concerted effort on understanding the broader issues surrounding why this demographic are leaving and arresting the trend could bring about a marked improvement in staff retention.

Whilst Staff Attrition should be one of the easiest metrics, in many instances this is not always the case.  What seems to make this more troublesome is the over complication of the basic measures.  Based on some intriguing rationales, the key measure of the staff population varies from a simple headcount, to the count of contracted Full Time Equivalence, or actual paid Full Time Equivalent for the period, the average of the Head Count by Month for a Year, average of actual paid Full Time Equivalent by Month for the Year, and so it goes on.  In all honesty, the key reason for using these convoluted definitions of Staff Counts is to manufacture a better result, and by better result, a lower Attrition Rate %age. When compared to using a simple headcount, the variance in the result gained by using an overly complicated calculation can be less than a percent.

There is considerable value in understanding and improving staff retention by using the Attrition Metric as an Analytic rather than a Performance Indicator.  Ultimately all staff are an investment, and the longer you can hold onto them the better the return on your investment.