Why Your Best People Leave — and Why You Found Out Too Late

He came into your cabin on a Tuesday and it took four minutes. He was calm, he had clearly rehearsed it, and he thanked you properly. You said the right things, asked what it would take, and heard a version of it's not about the money, sir, I've already committed.

What struck you afterwards was not that he left. It was that he had already finished deciding. There was nothing to discuss because the discussion had happened somewhere you could not see, over a period you cannot now reconstruct, and by the time it reached your desk it was an announcement.

Search this question and you will get five reasons, then seven reasons, then twelve. Bad manager. No growth. Low pay. Poor culture. They are not wrong, exactly. They are just the wrong unit of analysis, because they answer why and the thing that actually cost you this person was when.

A resignation is the fifth event in a sequence. The four before it were all visible, all cheaper, and all missed.

The number that changes how you read this

Start with the fact that most Indian promoters do not know.

Aon's India survey put attrition at 16.2% in 2025, down from 17.7% the year before and 18.7% in 2023. The market has cooled. That is the number people quote when they want to be reassured.

The number that matters is the composition. Nearly 75% of attrition in India is voluntary — against 50 to 66% in major global markets. That is Amit Otwani of Aon, and it deserves a moment. In most large economies, a meaningful share of departures are the company's decision: restructuring, performance exits, redundancy. In India, three-quarters of the people who leave your business are people who chose to. They were not asked to go. You did not plan for it. They simply decided, and the decision was theirs alone.

Which means Indian attrition is not a labour-market phenomenon happening to you. It is a series of individual decisions being made inside your building, one person at a time, and you are only told about the ones that reach the final stage.

Gallup's finding on that is the other half of the picture: 52% of voluntarily exiting employees say their manager or organisation could have done something to prevent them from leaving. Just over half of the people who walked out of your company would have stayed. Something specific, and by their own account not especially exotic, would have held them.

And the cost of not knowing which: Gallup puts replacing an employee at one-half to two times their annual salary, which for a senior person in an Indian mid-market business is a real number before you count the customer relationships, the six months of reduced output, and the two people who now wonder whether they should be looking as well.

The Exit Clock

Here is the sequence. It runs, in most cases, somewhere between nine and fourteen months from first stage to resignation.

Stage 1 — The Unanswered Ask

Somebody raised something real. Not a complaint — an ask. More scope. A decision they wanted to own. A title that matched the work they were already doing. A person they needed hired. Clarity on what came next for them.

You heard it. You may even have agreed with it. Then the quarter got busy, and it did not happen, and it was not refused either — it simply went quiet.

This is the cheapest moment in the entire sequence and almost nobody acts on it, because at this stage the person is still visibly committed. They are working hard. They are not sulking. There is no problem to solve, which is exactly why nothing gets solved.

The Culture Amp analysis of 175 teams is the evidence here, and it is worth sitting with. Across those teams, development opportunities drove 52% of departures. Leadership drove 28%. The direct manager, 12%. Pay, 11%. The single largest cause of people leaving is not being badly managed or badly paid. It is having nowhere to go.

Stage 2 — The Recalculation

This is where they actually leave. Everything after this is logistics.

Somewhere after the second or third unanswered ask, a capable person stops treating the job as a place they are building something and starts treating it as a place they are employed. The shift is internal and almost entirely invisible from your chair. They still deliver. They may deliver for another year. But they have stopped investing beyond the boundary of the role.

Gallup's India data describes the resulting population precisely: 59% of Indian employees are "not engaged" — present, compliant, doing the defined task, investing nothing past it. That is not a description of poor performers. In a promoter-led business it is very often a description of your solid, senior, long-tenured second line.

The reason exit interviews are useless is that they ask at stage five about a decision taken at stage two, by which point the person has constructed a tidy, polite, unfalsifiable narrative for it. We take that apart in why exit interviews don't work.

Stage 3 — The Quiet Look

Now they start looking, passively. Answering a recruiter they would previously have ignored. Updating a profile. Taking a call at lunch. Mentioning to one trusted colleague that they are open to hearing things.

This stage is long and it is the last one where retention is cheap. Nothing has been committed. No offer exists. The ask from stage one, granted now, would very likely still work — and would cost you a fraction of what the next two stages will.

It is also the stage with visible signals, and they are behavioural rather than emotional. Withdrawal from things that have no immediate deliverable attached. Stopping arguing with you. Declining the long project. We set out the six most reliable ones in employee flight risk indicators, because these are the signs a promoter can actually read without an HR analytics platform.

Gallup's job-search figures show how the population sorts at this stage: 74% of the actively disengaged, 55% of the not-engaged and 30% of the engaged are watching for or actively seeking a job. Note the last figure. Even among your genuinely engaged people, roughly three in ten have one eye open. Engagement is not a lock.

Stage 4 — The Offer

Now it is expensive. There is a number on a piece of paper, a start date, and — the part promoters consistently underestimate — a decision the person has already narrated to their spouse, their parents and at least one friend.

This is where the counter-offer happens, and where most Indian promoters make the same expensive mistake twice: they buy back a person who has already left internally, at a 30 to 40% premium, and lose them anyway in seven months, having reset the salary expectations of everyone who noticed. Should you counter-offer a resigning employee sets out the two questions that decide whether a counter-offer is ever the right call, and they are not the questions you would expect.

Stage 5 — The Resignation

The four minutes in your cabin. The only stage you can see, and the only stage at which you are trying to act.

That is the whole problem, stated as plainly as it can be.

Why the reasons on every other page mislead you

Two beliefs run this category, and both are more wrong than useful.

"People leave managers, not companies." This is the most-repeated sentence in the retention business. It is popularly traced to Gallup's research as told in First, Break All the Rules, and it has been compressed over twenty-five years into something the underlying work does not support. Researchers at Durham University Business School reviewed 39 papers from the past decade and concluded that people leave managers but also leave organisations — workload, pay, scheduling and progression clarity matter comparably, and even a good manager cannot retain people when the workload is high and the resources are not there. Culture Amp's 175-team data puts the direct manager at 12%. We deal with this properly, including who benefits commercially from the simplified version, in do people leave managers or companies.

"They left for money." Sometimes true, usually a summary rather than a cause. Gallup found 44% of employees would consider a job elsewhere for a raise of 20% or less — but that splits to 54% among the actively disengaged and 37% among the engaged. The same rupee moves a disengaged person and does not move an engaged one. Pay is the trigger, not the reason, and treating it as the reason is how a business ends up with a distorted salary structure and the same attrition. Do employees really leave for money works through where the pay threshold genuinely binds and where it does not.

How to run the Exit Clock on someone you cannot afford to lose

Pick the two or three people whose resignation would genuinely hurt. For each, answer four questions honestly, in writing, today.

• What did they last ask you for that has not happened? Not a complaint. An ask. If you cannot remember one, that is not evidence of contentment — it is evidence they stopped asking, which is stage two.
• When did they last argue with you about something that mattered? If the answer is more than four months ago, and they used to, something has changed.
• What is the next thing for them here — specifically, with a name and a rough timeframe? If you cannot say it in one sentence, they cannot either, and development is the single largest driver of departure in the data.
• If they resigned on Friday, what would you immediately offer? Whatever that is — the scope, the title, the number, the reporting line — you are willing to give it. You are simply waiting for a resignation to authorise it. Give it now and you will pay less for it.

That fourth question is the one that changes behaviour, and it is uncomfortable for a reason. Most retention is a timing failure, not a generosity failure.

Where to go deeper

The ten articles below each take one part of this apart. Start with the stage you are currently standing in.

• Do people leave managers or companies — the most-quoted claim in retention, traced to its source and tested against the evidence.
• Do employees really leave for money — the Pay Threshold, and the twenty percent line where a raise stops working.
• Employee flight risk indicators — the Six Withdrawals a promoter can read without any HR software.
• Why exit interviews don't work — the wrong instrument, pointed at the wrong stage, held by the wrong person.
• Should you counter-offer a resigning employee — the Counter-Offer Trap, and the only two questions that matter.
• Why high performers leave — the Utilisation Gap, and why your best person is the one most likely to be bored.
• Attrition in India: what the numbers actually say — the Aon data, sector by sector, and what the 75% voluntary share means for you.
• Quiet quitting in India — the internal resignation, and why 59% not-engaged is a bigger problem than 16% attrition.
• Why your managers leave — the Ceiling Test, and the second-line departure that costs a promoter-led business most.
• How to retain your key people — the Four Anchors, and what actually holds someone who has options.

Where to start

Do the four questions above, for two people, this week. It takes twenty minutes and it will tell you more than any engagement survey you commission this year.

If you want the structural version — where in your organisation people stop being able to grow, and which roles have no next step attached to them at all — the Business Pulse diagnostic maps it. Development opportunity is the largest single driver of departure in the evidence, and it is the one thing on that list you can build without spending a rupee more on salaries.