Executive Compensation: What Is Actually Negotiated

Compensation at a consequential seat is not settled as a number. It is settled as a structure: base pay, the architecture of the equity, the scope and decision rights attached to the seat, and the life terms around it. The number is one term in that structure, and it is rarely the term that decides the answer.

This page carries no benchmark table, and the reason belongs at the top. The club does not hold compensation data it can defend in public, and a figure that cannot be read, checked, and argued with is not worth printing. What can be written honestly today is everything around the numbers, which happens to be the part that actually moves in a negotiation.

Why this page prints no benchmark

A compensation benchmark is not a fact. It is the output of a method, and the method decides whether the number means anything at all. Change who was asked, the window they were asked about, or what counted as compensation, and the same market yields a different figure. None of that is visible in the figure itself, so a benchmark that arrives without its method is decoration, not evidence.

That method is a standard any reader can hold a publisher to, and it is the standard this page is held to as well. A figure is readable when it carries its sample size, its sample frame, the dates it was collected in, and a plain statement of what it cannot support; when cells thin enough to identify an individual are suppressed rather than printed; and when definitions stay fixed between editions, so one year can be compared with the last. Further down, that standard becomes five questions to ask out loud. No figure the club holds today answers them, which is why this page prints none.

The frame would matter as much as the method. The club’s mandates run in luxury and consumer houses, so any reading drawn from them would describe a defined segment rather than the market: a smaller claim than a market study, and a more defensible one. Every claim this Journal has made, with its source, is listed at the record, and a figure with no method behind it belongs nowhere near that page.

What a package is actually made of

Strip the theatre out of a senior offer and four things are being decided at once.

The number. Base pay is the most visible term and usually the one with the least room in it, because it is the term the house modelled, budgeted, and defended internally before the conversation started. The market’s own readers of pay say as much: Korn Ferry’s estimate for average salary increases in 2025 sits at 3.5% across all employee groups, with firms leaning on variable pay and retention instruments instead. The headline number is settled territory; the movement lives in the other three terms.

The equity structure. Not the headline figure but its architecture: which instrument, priced against what, vesting on what schedule, what happens to unvested value in a sale, in a down round, or in an exit that was not the holder’s choice, and whether the window to exercise outlives the employment. Two offers quoting the same equity value can be worth entirely different amounts once those terms are read. The reading, not the headline, is where a negotiation over equity actually happens.

The scope and the decision rights. Scope is what the seat is asked to return. Decision rights are what its holder controls in order to return it: budget, headcount, hiring authority, who reports in, and which of their decisions can be reversed above them. Both are written into an offer, and both are negotiable in the same conversation as the number. A title is a label. Decision rights are the job. An executive who carries a strong number into a seat with borrowed authority has bought a role they cannot perform, and the market will read the result as their failure rather than the structure’s.

The life terms. Where the work is done, how much travel, what happens when a family cannot move, what the first year asks of someone who already has a life. The club weighs location, family, lifestyle, and wellbeing in every match, on the standard that the work must fit the life and never a life bent around the work. A number that bends a life out of shape is not a win.

A number is one term in a contract. A package is the shape of the next five years.

The leverage an approached executive actually holds

The same package, negotiated by two people with identical records, does not settle in the same place. What differs is position.

An executive who applied is negotiating against a process. There is a requisition, an approved band, a shortlist, and a committee that priced the seat before meeting anyone in it. Their leverage is comparative, because the house’s alternative to agreeing is the next name on the list.

An executive approached about unlisted work is negotiating against a problem. Every mandate the club runs is exclusive to it, and the vast majority are never publicly listed, so there is no shortlist to reference. The mandate was opened around what the seat has to return rather than around a description of whoever last held it, the discipline walked in full in placing for the seat’s return. The house’s alternative to agreeing is not another candidate. It is another quarter with the work undone. That asymmetry is what moves equity terms and decision rights, and it belongs to someone who never raised a hand.

Two structural facts keep the position usable. Interest is reviewed privately and never shared with anyone, including a current employer, without explicit consent, so exploring costs nothing and leaves nothing behind. And every approach is confirmed with the principal first, so a conversation that starts has already been cleared by the person who can actually change the terms. How that position is held across a career is walked in negotiating from optionality, and the market’s own name for it in what a passive candidate is.

Where the club’s own numbers would have to come from

Two sources exist. Neither is ready to print.

The first is the room: 1,000+ members worldwide, peers drawn from the defining houses. That is a population worth asking, and a quarterly pulse of it is the honest way to read compensation from the inside, because it asks people who have just negotiated rather than people who publish ranges for a living. It becomes evidence only when enough of them answer the same question inside the same window, against definitions fixed before the first response arrives. Anything thinner is an anecdote with a chart on it.

The second is the placement record: 120+ executives placed into roles, each on terms that were agreed rather than advertised. Signed terms are the strongest compensation data there is and the hardest to publish, because they belong to identifiable people at identifiable houses. Anonymization has to come first, and it costs resolution. Cells standing on fewer than five placements are withheld rather than printed, and a thin cut is never dressed as a robust one.

Both sources carry the same limit as the frame named above: they would describe the rooms the market never lists, not the market. That is the smaller claim, and it is the one this page will make when it can.

How to read any benchmark you are handed

Before treating a compensation figure as a fact, ask it five questions. The list is worth sending to a peer with a table to sit down at.

1. Who was surveyed? A number is only ever about the population it was drawn from. Ask for the sample frame, the size, and who was left out. A study of companies willing to answer a compensation survey is a study of companies willing to answer a compensation survey.
2. When was it collected? A figure with no collection window has no shelf life. A report released this quarter may be reading decisions taken a year earlier, under conditions that no longer hold.
3. Self-reported, or verified? Ask what stands behind each figure: an offer letter, a payroll record, a recruiter’s file, or a memory. Where the answer is a memory, the number is an impression with a decimal point on it.
4. Base only, or total? A base figure and a total-compensation figure are different quantities, and a table that mixes them compares nothing. Ask what sits inside the number: bonus at target or at actual, equity at grant value or at an assumed outcome, and on whose vesting assumption.
5. What is suppressed? A careful publisher hides thin cells instead of printing them. If every cell in a table is populated, either the sample is unusually large or nothing is being suppressed, and a reader is entitled to know which.

A benchmark that answers all five is worth quoting. One that cannot is not evidence, whatever the logo on it, and an executive who carries it into a room has handed the other side of the table something to correct.

What the club will not do

No invented multiples. The club will not tell a member what they are worth as though it were a measurement, and will not repeat a figure it cannot source. A number handed to someone about to sit down at a table becomes either an instrument or a liability, and which one depends entirely on whether it is true.

No benchmark without a method. If the club ever prints a figure of its own, the method goes with it, ahead of the figures, whether or not they flatter anyone.

And no price on the member’s own access. Membership is free for talent, forever: the companies pay to reach the member, never the other way around, and the salary is the member’s own. What the club puts at the table instead of a benchmark is a person. The Talent Concierge handles negotiation and counsel for the length of every role process, complimentary: the same person who ran the process, who knows which terms the house can move, and who can say the direct thing so the member never spends goodwill saying it. Whether to answer the approach at all is a separate question, answered plainly in should you respond to a headhunter.

The standing offer is set out on the members page. Executives who would rather hold the stronger position before the next offer arrives can request an invitation, reviewed in confidence and visible to no one else.