Field notes. Opinions marked as opinions.

Attrition percentage tells you how many people left. It does not tell the CFO why he should care. For that you need a rupee figure, and the rupee figure is where HR reports tend to go wrong, in both directions. Some count only the advertisement cost. Others add up every conceivable loss and arrive at “twice annual salary”, a number nobody believes. Both get ignored.

Start with the arithmetic

Take a sales team of 400 with 35% annual attrition. That is 400 × 0.35 = 140 people leaving. Say you have already estimated that replacing one person costs ₹3 lakh. The annual bill is 140 × ₹3 lakh = ₹4.2 crore.

Now suppose a retention effort brings attrition down to 25%. Leavers become 100, and the bill is ₹3 crore. Saving: ₹1.2 crore a year. That is the whole calculation, two multiplications and a subtraction. The hard part is the ₹3 lakh. Spend your effort there, not on decorating the sums.

What goes into the per-leaver figure

Split it into things you can point to in the books and things you are estimating.

  • Separation: exit processing, final settlement admin, any notice-period shortfall, interviewer time. Small but real.
  • Replacement: agency fee (usually a percentage of annual CTC), job portal spend, referral bonus, recruiter time, manager interview hours, background verification, joining kit, induction.
  • Vacancy: work that did not get done, or got done on overtime or through a contractor, while the seat sat empty.
  • Ramp-up: the months before the newcomer performs like the person who left. In sales this shows up as lost quota.

The first two are hard costs. The last two are estimates. My opinion: report them separately. A leadership team will accept “hard cost ₹1.2 lakh, productivity loss estimated at ₹1.8 lakh” and argue only about the second figure. It will reject a single blended ₹3 lakh with no working shown.

Where the number lies to you

Not all attrition is equal. Losing 140 people is not 140 identical events. If 60 of them were low performers you were glad to see go, the true cost is lower, and the exit may even have been a gain. If 20 were your top closers, the true cost is higher than the average suggests, because their ramp-up loss is bigger and so is the damage to their clients.

So split the 35% into regretted and non-regretted exits before you multiply anything. Opinion again: a company that reports one attrition figure to its board is hiding the only interesting part.

Seniority matters too. The ₹3 lakh average is probably built from a mix of junior and mid-level exits. Replacing a regional head will not cost ₹3 lakh. Compute by band if your data allows it. If it does not, say clearly that the figure is an average.

Then there is the saving. The ₹1.2 crore is only real if the retention action costs less than that and actually causes the drop. Suppose you spend ₹40 lakh on a revised incentive scheme and attrition falls to 25%. Net gain is ₹80 lakh, provided the fall is yours. If the market cooled and everybody’s attrition dropped, you have claimed credit for the weather. Compare against a similar team that did not get the intervention, if there is one.

Things I would not include

Do not add a vague “loss of institutional knowledge” figure with a rupee value pulled from nowhere. If it matters, describe it in words and give an example. Do not multiply by a rule of thumb you picked up from a blog (you will see “50% to 200% of annual salary” quoted widely, usually without a source you can verify). Your own books beat anyone else’s ratio.

Also, do not forget the savings on the other side. While a seat is vacant, you are not paying that salary. A fair calculation nets this off against the vacancy cost. Skipping it overstates the loss and invites exactly the cynical question you do not want.

What to do with it

Put the number next to a decision. “Attrition costs us ₹4.2 crore a year, of which ₹2 crore is in the first twelve months of tenure” points at onboarding and hiring quality. “Attrition costs ₹4.2 crore, most of it regretted exits in two cities” points at a pay or manager problem in those cities. A cost figure with no decision attached is trivia.

And refresh it each year. Agency fees change, ramp-up periods change, and the first calculation is never the last one.

Figures and names in the examples are illustrative.

Leave a Reply

Trending