Unlocking the Power of Attrition Data: How Analyzing Employee Turnover Can Propel Organizational Success

In today's organizational landscape, understanding and effectively managing attrition data has become paramount for sustained success. Attrition and flow of talent within a company offer invaluable insights into workforce dynamics. From voluntary departures to internal transfers, attrition data encapsulates multifaceted reasons behind employee exits. By delving into this data goldmine, organizations can uncover patterns, anticipate trends, and strategically allocate resources. Let us jump straight into it.

What is Attrition and its types?

In simple words, an employee leaving a company is called attrition, we can add a hundred technicalities to this but that is not the agenda for the day. The thing that is actually important to understand in this article is the types of attrition.

Again, we can have as many classifications under the banner of attrition but that doesn't solve the purpose of the day.

Involuntary Attrition

Involuntary attrition refers to the process of employees leaving their jobs due to reasons beyond their control, such as layoffs, terminations, or restructuring within the organization. Involuntary can include anything where the will of the employee was not taken into consideration.

Voluntary Attrition

Voluntary attrition is when employees choose to leave their jobs voluntarily, such as resigning to pursue other opportunities or retiring. This kind of attrition is usually studied in more depth by considering whether the employee leaving the organisation is to a competitor or due to personal reasons of employees, hence to make following sound decisions.

Nevertheless, now as we understand things on a basic level, let us move on to the agenda of the day.

Level 1 Attrition Analysis: Attrition Dashboard

We call level 1 analysis, something that is done on the face value of the data without involving any further combinations. As an example of such analysis, we have attrition dashboards answering 4 questions:

Who is leaving the organisation?

When did they leave?

Where did they go?

Why did they leave?

Let us try to understand the real idea behind each of these questions.

Who is Leaving the organisation?

We are trying to solve, the person is leaving from which department, which sub-department, which PnL or which business group. Is one group facing more attrition than others? Have any recent changes been leading to a rise in these specifics? Usually, attrition remains high in client-facing roles, as in the sales force for a sales-first organisation.

One of the most important metrics under the "Who" header is manager-wise attrition. Try to study this with respect to performance data, you will amazed to see the end result. Try checking out, Why are employees underperforming & how do we solve it?

When did they leave?

Whether the attrition that we talking about happen in the first few months or after 2-3 years? What is the average span that an employee is spending in our organisation? Whether our organisation is being used as a launchpad for the industry and later they move on to other organisations, in short, we might need to change our strategy. Is there an element of seasonality that the attrition happens in a particular month?

Where did they go?

It is very important to track this metric and this data point has the least available possibility in an organisation. Whether an employee from our particular department is leaving us for a competitor? This gives us the target group to focus on while we are looking for a policy change.

Why did they leave?

Consider a situation where an employee is going to leave an organisation. On a scale of one to ten, how much would you rate the employee's honesty in answering organisation-related questions?

What would you use to gather real data? Is only survey the right choice? Even if you are going for a combined choice of survey and interview, are the questions in your survey set in the right format to get real insights? Let us assume you have the right skill to set the question in the best format possible, are you using interviews to verify the claims of the survey or are you double downing on the same data points? Tough calls right?

In my experience of attrition data collection, organisations aren't actually doing a very good job. Just setting up the processes in the same format or the same question as done by someone else in the industry, isn't going to solve your problems. Let me know, if I should be going deep on the topic of how to set attrition process in an organisation.

Or we can always have one on one connects. Raghav Rozra

The problem with most organisations is that they don't even have these basics clear in their dashboards.

Level 2 Attrition Analysis: Attrition Dashboard

We have almost reached mid of the article, congratulation, you are no more a dashboard noob, in other words, now you understand what you are doing.

But let us move a step ahead,

For the sake of this article let's consider 2 data connections:

1. Attrition and PMS data
2. Attrition and Talent Acquisition data

Attrition and PMS data

What does this even mean? Why should we study these data points together? Is it even worth the effort?

Both our attrition and PMS data have employee IDs mapped to each of our employees, hence connecting isn't an issue at all.

But why?

So that we can study whether we are losing our best employees to competition or not. To decide whether we need to worry about the manager under whom the attrition percentage has never been below 40%. To decide whether a cohort of employees is getting targeted by our competitors. For example, As we discussed in our How to manage the performance of ground-level sales employees? blog, assume our Performance Management System is burdened under the loop of force-fitting, employees doing everything well yet scoring 4 and not able to get the promised appraisal subsequently leave the organisation. If this data is well combined with the initial data point of "where?", we will be able to catch the problem in a much earlier stage.

Attrition and Talent Acquisition data

Now consider the idea that after many discussions, the HR TA team decided on six companies as their target groups for hiring a particular level of employees. Turns out that several employees are leaving from that level. How would you know what to do?

Rather consider the TA and HRBPs (dealing with attrition) have no strategic discussions around this particular problem as the TA team got its approvals from senior management and hence have started focusing on hiring more to compensate for losing employees. This is the point when your attrition data linked with the target groups selected by the TA team, will itself show the blunder that the organisation had made over time. As simple and as clear as saying, 78% of hiring from company X, leave us within 6 months. Now, it is the organisation's call whether you wish to push in some retention strategies or change your target groups, but everything starts with having that golden data in hand.

Over and above the examples given above, an organisation can have tens of use cases depending on your business cases and needs. The thing that I am trying to stress here is, attrition data is an amazing bullet to have if used with the correct pair of connected triggers.

Predictive Attrition Analytics

In the era of the talks of AI, let me ask you an important question.

How many organisation around you have an actual trustworthy attrition prediction in place?

I may be wrong here, but not even 10% is what I will place my bet on. By prediction, I am talking about a system that is capable of answering most of our future attrition-related questions.

For this article let's just try and imagine the endless possibilities this system holds if we are able to connect the predictive attrition employees to their TGs, their managers, their departments or an early system of predicting that is able to give us the data that the employee of this particular cadre or sub-department is gradually losing interest. Rather this is not a very far reality. As I always say, before impactful insights, it is the targeted measuring of data that matters. If we are measuring the right data and using it the right way, we may very well be able to do this in the next 6 months. My preferred tool for such predictive analysis will always be R, with endless possibilities to explore in the most open source possible.

Now for the RIGHT measuring and RIGHT data, let me reserve the right to that information for the day. Let me know if you would like to have an article dedicated to this topic.

Have a nice day and let me know your views on the same. Let us explore some unsaid possibilities of such analysis in the comments.

Would love to hear your views on the same!

Subscribe to the newsletter for latest updates