Attrition Rate in IT Results: Reading Headcount, Utilisation and TTM Attrition Together

Every quarter, India’s IT companies publish one small line that tells you more about their next two quarters than most of the management commentary. It’s the attrition rate in IT results: the share of employees who left over the last twelve months. On its own it’s a people statistic. Put it next to headcount and utilisation, and it starts telling you whether a company is short of staff, sitting on a fat bench, or quietly shrinking to protect margins.

This guide covers what the number actually measures, why companies report it on a trailing basis, and how we read it alongside the other two people metrics. We’ll use the September 2026 quarter disclosures from Wipro, LTM and Tech Mahindra as real examples, and a made-up company for the arithmetic.

What the attrition rate in IT results measures

At its simplest, attrition is the rate at which people leave a company over a period. Investopedia’s definition gives the basic formula: number of departures divided by the average number of employees, multiplied by 100. The average is the headcount at the start of the period plus the headcount at the end, divided by two.

Indian IT companies almost always report this on a trailing twelve-month (TTM) basis. Instead of counting exits in one quarter, they count exits over the last four quarters and divide by average headcount over that year. That smooths out the quarter-to-quarter swings that would otherwise make the number jump around.

The catch is that every company defines the number its own way. Read the label before you compare anything:

• Wipro reports voluntary TTM attrition for IT services, excluding its business process services (BPS) arm. BPS gets a separate quarterly figure.
• LTM, the Larsen & Toubro group IT company, reports a single TTM attrition figure in its investor release.
• Tech Mahindra reports last-twelve-month IT attrition.

“Voluntary” means people who resigned. It leaves out people the company let go. So a company can cut staff and still show a falling voluntary attrition rate. Keep that in mind; it matters in the Wipro example further down.

Worked example: calculating TTM attrition

Let’s take Kaveri Infotech, a hypothetical mid-size IT company. Every figure for it in this article is invented purely to show the arithmetic.

• Headcount on October 1 last year: 40,000
• Headcount on September 30 this year: 44,000
• Voluntary exits over those twelve months: 5,880

Average headcount is (40,000 + 44,000) / 2 = 42,000. TTM attrition is 5,880 / 42,000 × 100 = 14.0%.

Now notice what happened. Kaveri lost 5,880 people and still ended the year 4,000 bigger. That means it hired roughly 9,880 people in gross terms (4,000 net growth plus 5,880 replacements), ignoring involuntary exits. A 14% attrition rate isn’t a crisis for a company growing like that. The same rate at a company with flat revenue means a lot of hiring just to stand still, and every replacement hire costs money. Investopedia notes that the cost of hiring and training a replacement can run from half to two times the departing employee’s annual salary.

Attrition tells you how many people walked out. Headcount tells you whether the company bothered to replace them. You need both before the number means anything.

Headcount: the net number hides the gross churn

Headcount is the closing employee count at the end of the quarter. Analysts watch the net change: how many people the company added or lost compared with the previous quarter. But net change is gross hiring minus all exits, and a small net figure can hide a lot of movement underneath.

Back to Kaveri. Say it starts a quarter with 44,000 people and its 14% annual attrition runs at about a quarter of that pace in three months. That’s roughly 44,000 × 14% / 4 = 1,540 exits. If it hires 1,000 people in the quarter, headcount ends at 43,460. Net headcount fell by 540, even though the company hired a thousand people. Exits don’t arrive in neat quarterly slices in real life, but the direction holds.

This is why we never read a headcount drop as “layoffs” or a headcount rise as “strong demand” without checking the other two numbers first.

Utilisation: how much of the workforce is actually billing

Utilisation is the share of employees who are billed to clients. Wipro and LTM both quote it excluding trainees, because freshers in training can’t be billed anyway. People who aren’t billed are “on the bench”.

Kaveri again. Suppose 38,000 of its employees are billable (excluding trainees) and 32,300 of them are working on client projects. Utilisation is 32,300 / 38,000 = 85.0%. If the company pushes that to 87% on the same base, it has 33,060 people billing, 760 more, without hiring anyone.

That’s the lever IT managements pull when demand is uncertain. Higher utilisation lifts margins because the company is paying the same salaries and billing more of them. But it can’t go up forever. When utilisation is already high and attrition starts rising, the company has to hire fast, often at higher salaries, and margins take the hit.

Reading the three numbers together

Here’s how we combine them when a results release lands:

• Attrition falling, headcount falling, utilisation flat: the company isn’t replacing everyone who leaves. Usually a sign it sees soft demand and is letting natural exits trim costs.
• Attrition falling, headcount rising, utilisation falling: it’s hiring ahead of expected work and the bench is growing. Fine if deal wins back it up; a margin drag if they don’t.
• Attrition rising, utilisation high: the classic pressure point. Expect replacement hiring, wage costs and possibly a margin squeeze in the next couple of quarters.
• Attrition stable, headcount and revenue both growing modestly: the boring, healthy case.

Then we check those signals against revenue growth and management guidance. People numbers are an early read on what management expects. Revenue is what actually happened.

Real example 1: Wipro, September 2026 quarter

Wipro’s Q2 FY27 datasheet gives all three metrics with history:

• Closing employee count: 236,328 at the end of September 2026, against 243,044 at the end of June 2026 and 235,492 a year earlier. That’s 6,716 fewer people than the previous quarter, but 836 more than a year ago.
• Net utilisation (excluding trainees): 83.4%, against 83.6% in the June quarter and 86.4% a year earlier.
• Voluntary TTM attrition (IT services, excluding BPS): 13.8%, against 13.9% in the June quarter and 14.9% a year earlier.
• BPS post-training quarterly attrition: 10.7%, up from 9.9% in the June quarter.

Read that against the results press release: IT services revenue was flat sequentially in constant currency and up 0.6% year-on-year, and Wipro guided for sequential constant-currency growth of -1.5% to +0.5% in the December quarter.

So voluntary attrition barely moved, yet the employee count dropped by 6,716 in three months. Utilisation was close to flat quarter-on-quarter and 3.0 percentage points lower than a year ago. The datasheet doesn’t say why headcount fell. It could be fewer replacement hires, exits that the voluntary IT number doesn’t capture, or movement in BPS, and we’d rather not guess. What the numbers do show is a company bringing its workforce in line with flat revenue and cautious guidance. We covered the quarter in more detail in our Wipro Q2 results report.

Real example 2: LTM, September 2026 quarter

LTM’s Q2 FY27 investor release filed with NSE shows a steadier picture:

• Total employees: 88,835, up 949 from 87,886 in the June quarter and up 2,388 from 86,447 a year earlier.
• TTM attrition: 13.3%, the same as in each of the previous two quarters, and 90 basis points lower than the 14.2% of a year earlier.
• Utilisation (excluding trainees): 86.4%, unchanged from the June quarter and 170 basis points lower than the 88.1% of a year earlier.

Revenue grew 1.1% quarter-on-quarter and 5.1% year-on-year in constant currency, according to the same release. This is closer to the “boring, healthy” pattern: modest hiring, stable attrition, modest growth. Utilisation is lower than a year ago, which means LTM has some spare capacity to take on new work before it needs to hire aggressively. That’s a small cushion for margins if deal ramp-ups come through.

When utilisation is lower than last year and attrition is stable, a company can grow for a while without a hiring spree. When utilisation is already stretched, every new deal comes with a wage bill attached.

Real example 3: Tech Mahindra, and why definitions matter

Tech Mahindra’s Q2 FY27 press release reports total headcount of 152,206 and last-twelve-month IT attrition of 11.7%.

Line that up with the other two and Tech Mahindra’s 11.7% sits below LTM’s 13.3% and Wipro’s 13.8%. It’s tempting to rank them. Don’t push it too far. Wipro’s figure is voluntary attrition for IT services excluding BPS. Tech Mahindra’s is IT attrition. LTM’s is a company-wide TTM figure. They are close cousins, not the same measurement. The cleaner comparison is each company against its own history: is the number rising or falling, and how fast?

Common mistakes when reading attrition

• Treating a falling attrition rate as purely good news. It often is. But in a weak hiring market, people simply have fewer places to go. Low attrition plus falling headcount plus soft guidance is a demand story, not a culture story.
• Comparing single-quarter attrition with TTM attrition. Wipro’s BPS number is quarterly; its IT services number is trailing twelve months. They aren’t comparable to each other, let alone to another company.
• Ignoring the trainee line. A big fresher intake pushes headcount up and doesn’t show in utilisation excluding trainees. That’s a bet on future demand, and it costs money today.
• Reading one quarter in isolation. Wipro’s datasheet and LTM’s investor release both give five or more quarters of history. Use it. The trend over four or five quarters tells you far more than one print.

A quick checklist for the next IT results release

1. Find the attrition number and its exact definition: voluntary or total, IT only or company-wide, TTM or quarterly.
2. Compare it with the last four quarters of the same company.
3. Check the net headcount change and work out whether the company is replacing its leavers.
4. Look at utilisation excluding trainees: is there bench capacity, or is the workforce stretched?
5. Line all three up against constant-currency revenue growth and the next-quarter guidance.
6. Then read the employee cost line in the profit and loss statement to see what it’s all costing.

That last step is where people metrics turn into margins. Our guide to reading employee cost in quarterly results picks up exactly there, and for the full framework of what to check in any earnings release, see our guide on how to analyse quarterly results of a company.

Disclaimer: This article is for educational purposes only and is not investment advice. Kaveri Infotech is a hypothetical company, and every figure attributed to it is made up for illustration. Company figures for Wipro, LTM and Tech Mahindra are taken from their September 2026 quarter disclosures linked above; the changes and differences were calculated by us. Stock investments carry risk. Please do your own research and consult a SEBI-registered investment adviser before making any investment decision.