Do Employees Really Leave for Money?
He got thirty percent, sir. What was I supposed to do?
You have said this, or heard your HR head say it, and the room accepted it. Nobody asked a follow-up question. Another company paid more, the market is the market, and the conversation moved on to the replacement.
It is an efficient explanation. It is also, usually, a description of the last week of a decision that took eleven months.
The claim this article makes is not a soft one. Pay is almost always the trigger and almost never the cause. That distinction sounds like wordplay until you notice what it decides: whether the money you are about to spend will work. Buy the trigger and the cause stays where it was, now with a larger salary attached to it.
Culture Amp analysed 175 teams across hundreds of fast-growing companies and found development opportunities drove 52% of departures, leadership 28%, the direct manager 12%, and pay 11%. Pay came last. It is real — 11% is not zero, and we will come to exactly who those people are — but it is the smallest of the four.
The resignation you are reacting to is stage five of a sequence we set out in why your best people leave. The offer letter arrives at stage four. The person left, internally, somewhere around month three.
Why "he left for money" is the story everyone agrees on
A version of events that fits the data this poorly survives because nobody is lying and nobody has a reason to correct it.
The employee needs it. He is on a ninety-day notice period, in a country where reference calls happen informally between promoters who know each other. I got a better package is unarguable and closes the exit conversation in four minutes. My manager takes credit for my work and I have not learnt anything since 2023 requires a conversation he does not want to have with a man he will need on a phone call one day. He picks the safe sentence. Anyone would.
The manager needs it. If the man left for money, the manager did not lose him. It is a market event, like a competitor cutting prices. The alternative version puts the manager's name in the explanation, and he is the one writing the handover note and briefing you.
And you need it most. A market explanation is external and beyond your control, which is oddly restful. An internal explanation means something in your business is producing this, that it will produce it again next quarter, and that fixing it involves you. Given the choice between a problem you cannot be blamed for and one you must personally act on, most promoters take the first without noticing they chose.
Three people, three incentives, one agreed sentence. That is why it never gets corrected — and why your exit data is worthless and your salary bands keep moving.
The Pay Threshold
Pay does not behave in one way. It behaves in three, depending on where the person sits relative to what their role pays in your city. Most arguments about compensation are two people describing different positions on this line, each assuming the other is wrong.
Below the threshold, pay is a cause
If someone is materially underpaid against the market for the work they do, nothing else you build will hold them. Not culture, not values, not the fact that you were at his wedding. A man doing a sum on his phone about school fees will not be talked out of it by purpose. This is the 11%, and for these people pay is not the trigger. It is the whole reason.
Indian mid-market businesses sit here far more often than they believe, and almost always with their most loyal staff. The mechanism is arithmetic, not intent.
Your internal increment is a percentage of what the person already earns. The market rate is a fresh price for the role, set each year by whoever is hiring hardest for it. Two different calculations, compounding at different speeds. Run them side by side for a decade and they separate — quietly, annually, with nobody deciding anything.
Which produces the situation you can verify this afternoon: the fifteen-year employee is frequently the worst-paid person doing that job in the city. He is your most trusted man. He has never asked, and he will not, because after this many years asking feels like a threat and he has built an identity around being the one who does not. When he finally goes, he goes for a number that will shock you — not because the market is mad, but because you had been paying below it for six years and calling it loyalty.
Take your five longest-tenured people and find out what it would cost to hire each of them today, from outside. Not what you think — what a consultant would quote you. The gap is your exposure, and it is a number you owe them rather than a negotiation you are losing.
At the threshold, pay is a tiebreaker
Second position. The person is paid roughly correctly, has two options that are genuinely close on everything else — scope, commute, boss, brand — and money breaks the tie.
This is the only place on the line where a raise does exactly what you hoped. It is decisive, fast and cheap, because at the threshold the gap you need to close is small. A person who is otherwise happy is not looking for a large correction. He is looking for a reason to stay that he can explain to himself.
If you are going to spend money on retention, spend it here, and spend it before anyone resigns. Almost nobody does, because there is no trigger to prompt it. Nothing forces the conversation. That is precisely the point.
Above the threshold, pay is a trigger only
Third position, and the expensive one. The person is paid fairly. He is not leaving because of the money, but the money makes leaving executable — it converts a vague dissatisfaction into a decision with a date on it, and supplies the explanation everyone will accept.
Pay him more here and you buy months, not years. You have addressed the mechanism of departure and left the reason for it intact. This is the population that resigns again two quarters after a counter-offer, which we take apart in should you counter-offer a resigning employee.
Gallup draws the line between the second and third positions sharply. 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. Same offer, same rupees. It moves one man and not the other, and the difference has nothing to do with compensation. The money is not the variable. The state they are already in is the variable.
The Twenty Percent Line
That split gives you something more useful than a benchmark. It gives you a measuring instrument.
The size of offer it takes to move your people tells you how engaged they are.
If your good people are leaving for a 10 to 15% increment, stop discussing salary bands. Nobody with an interesting job, a boss he respects and a visible next step changes company, commute, provident fund, notice period and the entire social structure of his working day for 12%. The maths does not work for him and he knows it. If he is doing it anyway, the increment is not the payment — it is the permission.
You do not have a pay problem. You have an engagement problem being expressed in rupees, and paying it will not fix it.
If your people are staying through offers of 25 or 30%, the reverse holds. That is not loyalty and it is not your brand — it is engagement, and it is buying you good people below the market clearing price. That is an asset with a rupee value, and it depreciates quietly. Employee flight risk indicators covers the behaviour that shows up when it starts to.
How to test it
This takes an afternoon and it is uncomfortable in a useful way.
Take your last three departures — the ones where money was the stated reason. Find out what each is earning now. Somebody in your business knows, or knows the person who knows; salaries travel through WhatsApp faster than any other information in this country. If you kept a decent relationship, ask the man directly. Most will tell you once they are safely out.
Then look at the increase.
If the increase was under about 15%, money was the stated reason and it was not the actual one. Nobody changes company, boss, commute and daily routine for 12% when things are good where they are. The number was the exit door, not the destination. Something else pushed him through it, and it is still in your business, doing the same thing to the next person.
If the increase was very large, you have the opposite finding and it is equally useful: you were below the threshold, and the market corrected you. Go and check who else is doing that job at that salary before you receive their resignation too.
Three people. One afternoon. It will tell you more than the benchmarking exercise you are being quoted for.
The salary-band damage
Now the part that gets underestimated, because its cost arrives late and in a different account.
You pay a retention premium. A senior person resigns, you go well above his current salary to hold him, and it works — he stays. You consider the matter closed.
The number is known inside about ten days. Not because he announced it, though sometimes he does — because salary information in an Indian business moves through the same channels as everything else: a car ride to a client meeting, a colleague in accounts, one WhatsApp group. It has never once been confidential.
Two things then happen, both invisible on your P&L this quarter.
The two people at his level recalibrate. They are not aggrieved on principle — they are simply revising their own number upward, and from now on every increment you give them is measured against his, not against last year's. You have reset a reference point for a whole layer.
And the organisation has learnt something durable: the reliable way to get a raise here is to produce an offer letter. Nobody says it. Everybody now knows it. Over the next eighteen months the good ones will start taking recruiter calls not because they want to leave but because that is what the process requires — and some of them will find, while doing so, that they do.
This is the single most expensive habit in Indian mid-market businesses, and it never appears as a line item. You have converted your compensation system into an auction in which you only bid after someone else has. The other version of this misdiagnosis — putting the entire cause on the manager — is dealt with in do people leave managers or companies.
When to pay, without apology
None of the above argues for being tight. There are three cases where you should pay, and pay properly, and the only real error is paying late.
Genuine market correction. You have run the numbers and this person is meaningfully below what the role commands. Pay it. Do not stage it over three years. Do not treat it as a favour, because it is not one — you were under-paying, whether or not you meant to.
Scope that has grown without a review. The role he is doing today is not the role he was hired for. He absorbed a function when somebody left, or the business doubled around him and he carried it. His title and his salary describe an old job. This produces the angriest departures, because the man feels he was quietly charged for his own competence. Why high performers leave is largely a description of this.
You cannot replace them this year. Sometimes this person holds a client relationship, a plant licence or institutional knowledge you cannot rebuild in twelve months. Paying above your band is not weakness, it is a correct commercial decision — provided you know you are buying time and you use that time to build the redundancy.
In all three cases, pay before the resignation, not after. The same money spent six months earlier costs less, holds longer, and teaches the organisation something entirely different about how raises happen here.
Where to start
Do the three-departures test this week. Find the actual increments. If they cluster under 15%, you have your answer, and it is not a compensation answer — how to retain your key people sets out what actually holds someone who has options.
If you want the structural view — where your bands have drifted below market, which long-tenured people are exposed, and which roles have no next step attached to them at all — the Business Pulse diagnostic maps it. Most promoters find both problems in the same business at once, in different people, having treated all of it as one pay problem.