How compensation is really decided — bands, levelling, internal equity and the approval chain — and what that means for the negotiation you can actually win. How to read every line of an offer, including private-company equity that may be worth nothing, with literal scripts for countering, asking for time, handling a best and final, and accepting.
Masterclass·88 min read
What it is
Negotiating an offer is the short window between someone deciding they want to hire you and the contract being signed, during which the price of the arrangement is settled. It is the only part of a job search where the balance of information and interest has moved in your direction, and it is usually the part candidates have thought about least, because everything up to that point felt like the hard bit.
The window is narrow and it has a shape. It opens the moment a verbal offer is made, or slightly earlier if the recruiter starts sounding you out on numbers before the panel has finished deliberating. It closes when you sign. Inside it, a handful of things are genuinely movable, several things are movable only by someone you will never speak to, and a few things are fixed by policy and will not move for anybody. Most bad negotiations are not caused by weak nerve. They are caused by spending your one credible ask on something that was never going to move, or by arguing hard with a person who has no authority to agree.
The other thing worth saying at the outset is that most writing on this subject overstates your leverage. It is written as though every candidate is a scarce senior engineer with three competing offers in a hot market, and it advises accordingly. Plenty of candidates are negotiating from a much weaker position: one offer, a notice period running out, a redundancy behind them, a visa clock, a market that is not hiring. That position is not hopeless and it is not the same position, and pretending otherwise produces advice that gets people hurt. This page tries to say plainly, at each point, where you have room and where you do not.
The path from the first salary question to a signed contract has one branch that determines how much of the rest matters.
Loading diagram…
Look at the branch out of the counter. By the time the verbal offer arrives, the level has usually already been set during the loop, and the level determines which band the number comes from. If you are in the band you wanted, the counter is a negotiation about a number. If you are in a band below the one you wanted, the counter is a negotiation about the number inside a container that has already been chosen for you, and the ceiling is much lower than it looks.
The parties, and what each of them wants
A negotiation reads very differently once you know who is on the other end of it and what they are being measured on.
The recruiter is your counterparty in almost every conversation and is the person with the least authority in the process. Their job is to close you, at a number the company is comfortable with, without losing you and without creating a problem for the team you are joining. They are frequently measured on offer-acceptance rate, which means they have a real interest in you saying yes, and sometimes on time-to-fill, which means they have a real interest in you saying yes quickly. Crucially, they are usually not the person who sets the number and often not the person who can approve a change to it. They can decline on the company's behalf without asking anybody. They cannot usually agree to anything unusual without asking somebody.
The hiring manager wants you specifically, has spent weeks of their own time getting to this point, and does not want to restart the search. They have views about what the role is worth and they usually have a budget position, though the budget is rarely theirs alone. They are the person whose enthusiasm most reliably converts into movement, because they are the one who will go and argue for an exception. They are also the person you will report to, which is the reason to keep every exchange courteous even when it is firm.
The compensation or reward function owns the bands, owns internal consistency, and is structurally sceptical of exceptions. They are not being obstructive for its own sake. They are protecting a pay structure that gets audited, benchmarked and occasionally litigated, and an out-of-band offer to one candidate creates a documented anomaly that somebody will have to explain later. When a recruiter says "that is above the band and I would have to get an exception approved", that is usually true rather than theatre.
The finance or headcount owner cares about the annual cost of the role against a plan agreed some time ago. This is the party most often behind an inflexible base and a flexible one-off payment, because a base increase recurs forever and compounds through every future rise, while a sign-on bonus lands in one year's budget and disappears.
Understanding this cast changes what you ask for. A request the recruiter can grant unilaterally gets answered in an hour. A request that requires a levelling change requires a different set of people and a different sort of evidence, and it is not a conversation you can win by being firm on a phone call.
The early salary question, and why it is not this page
Most processes ask what you are looking for long before an offer exists, usually in the screening call. That question is a different problem with a different answer, and it belongs to what are your salary expectations and to the recruiter screen guide rather than here.
The one-line version, because it affects everything downstream: try once to have them name the budgeted range, because they have one and you do not; if pressed, give a researched range tied to the market for the role rather than to what you currently earn; and do not treat whatever you say as binding, because it is not. In many places there are rules about asking for salary history, the shape of which is roughly that employers may not require it and must not use it to set your pay, but the detail varies a great deal by jurisdiction and you should not assume any particular protection applies to you. Whether or not a rule protects you, the practical advice is the same: your current pay is a fact about your last employer, not about this job, and offering it hands the other side a free anchor.
The number you gave at screening matters here in exactly one way. If the written offer comes in below it, that gap is the cleanest thing in the world to raise, and the conversation is already framed for you. That specific situation is covered at the written offer is below the number you gave.
The case for negotiating at all is not the one usually made. It is not that companies are trying to underpay you, although some are. It is that a first offer is a proposal made under uncertainty, generated from a band, a levelling decision and a guess about what you will accept, and proposals made that way are frequently not at the top of what the organisation would pay. Nobody is being dishonest when they open below their ceiling. They are opening at a number that is comfortable, because that is what a first offer is for.
The second reason is structural and compounds. Your starting salary at an employer is the base from which every subsequent rise is calculated, and most internal rises are expressed as percentages of it. Two people hired into the same role at different points in the same band tend to stay in the same relative position for years, because the mechanism that would correct it — an out-of-cycle adjustment — requires someone to notice and to expend political capital. The gap you accept on day one is not a one-year gap.
The third reason is that the offer stage is the only point in the relationship where you have an alternative that costs the other side something. Once you have resigned, once you have started, once you have moved cities, the alternative is gone and the leverage with it. Whatever you want from the arrangement that is unusual — the remote days, the start date, the title, the review at six months — this is when it is cheap and afterwards it is expensive.
What the employer is protecting
It helps to know what a "no" is usually defending, because the defence tells you whether there is a different route to the same thing.
They are protecting internal equity, which is the single most common real reason a number will not move. If the person doing your job three desks away is paid a certain amount, and you are hired above them with the same title and less tenure, that is a problem waiting to detonate the first time pay is discussed over a drink. Companies that have been burnt by this are rigid about it, and it is the reason a recruiter will sometimes tell you plainly that they cannot go above a figure that is obviously not a hard band ceiling.
They are protecting the band's integrity, because a band with routine exceptions is not a band. Levelling frameworks only do their job if the mapping from level to pay holds, and every exception weakens the argument the next time someone else asks.
They are protecting the precedent. A recruiter who grants an unusual concession has to grant it again to the next person who asks, or explain why not. This is why a request framed as a general policy question ("do you ever do that?") gets a harder no than the same request framed as specific to your circumstances.
They are protecting the recurring cost. A base increase is permanent and multiplies through pension contributions, bonus percentages and every future rise. A one-off payment does none of that. This asymmetry is the single most useful thing to know in the whole subject, because it tells you which of your asks is expensive and which is cheap, and lets you take the money in the shape they can afford to give it.
Where your leverage genuinely is, and where it is not
Be honest with yourself about which of these you have, because the strategy differs.
You have real leverage when you are the only finalist and the search has been long, when the role has a deadline attached to it such as a project start or a departure being backfilled, when you have a second offer that is genuinely comparable, or when you are currently employed and visibly not desperate. In those situations a firm, well-justified counter has a good chance and a small downside.
You have modest leverage when you are one of two acceptable candidates, when the market for your skill is soft, or when you have been out of work for a while and they know it. Here a single reasonable counter is still worth making, phrased so that it cannot be read as a condition. The expected value is positive and the risk, handled well, is close to zero.
You have little leverage when you need this job on a timeline you do not control — a notice period expiring, a visa or right-to-work clock, redundancy pay running out, a relocation already committed. That situation deserves its own treatment and gets one in the job hunting after redundancy guide, which is about negotiating from exactly this position. What is worth saying here is that little leverage is not no leverage, and the non-monetary asks later on this page are often available even when the number is not.
The mistake in both directions is symmetrical. Candidates with leverage frequently do not use it, because asking feels rude. Candidates without leverage sometimes negotiate as though they had it, because an article told them to, and the cost is real.
How compensation decisions are really made
Most candidates negotiate against the recruiter as though the recruiter is choosing the number. They are not. They are reporting a number that came out of a machine, and the machine has parts.
Bands and levels
Nearly every organisation past a certain size maps roles onto levels, and each level onto a pay band with a floor, a midpoint and a ceiling. The band is wide enough to accommodate a range of experience within the level and narrow enough that two people at the same level are not paid wildly differently. Where in the band a new hire lands depends on their experience relative to the level, on what the internal population at that level currently earns, and to some degree on what the recruiter thinks it will take.
Three consequences follow, and they are the whole game.
First, the ceiling of your band is a real ceiling and a counter above it is not a negotiation, it is a request for an exception. Exceptions exist but they are rare, they require a named approver, and they are usually reserved for a candidate the hiring manager is willing to spend capital on.
Second, where you sit within the band is genuinely negotiable and is where most successful negotiations happen. A first offer is often placed below the midpoint precisely to leave room, particularly for a candidate who is new to the level.
Third, and least appreciated: the band you are negotiating inside was chosen when your level was decided, and that happened before the offer conversation started.
Internal equity, and why it produces strange answers
Internal equity is the constraint that makes an otherwise inexplicable "no" make sense. If the company has four people at your level in your function, their pay forms a distribution, and a new hire placed above the top of that distribution creates a problem. The problem is not abstract. It surfaces at the next pay review, when a five-year employee discovers the new starter is paid more, and the manager has to either correct it, which costs money that was not budgeted, or not correct it, which costs a person.
This is why a recruiter will sometimes hold firm at a number that is visibly below the band ceiling. They are not at the band's limit. They are at the limit of what they can pay you without creating an internal problem, and that limit is invisible to you and non-negotiable in principle. If you suspect this is what is happening, the tell is a "no" that is confident, immediate and does not come back with a compromise. A budget "no" usually comes back with something.
Budget, approval and who can actually say yes
The approval chain is short but real, and knowing its shape tells you how long a counter should take and what a long silence means.
Loading diagram…
The useful thing here is the timing signature of each path. A counter that stays inside the band and inside the hiring manager's discretion comes back within a day or two. A counter that needs the reward team comes back in three to five working days and often comes back partially granted. A request that reopens the levelling decision takes longer than either and frequently comes back as a no with an explanation, because reopening it means telling a panel their assessment was wrong.
So if you counter and hear nothing for a week, that is not usually a bad sign and it is not usually the recruiter ignoring you. It means your ask left the recruiter's authority. Chasing politely once after three or four working days is correct and costs nothing.
Which components move, and who has to agree
Component
How negotiable it usually is
Who approves it
Base salary within the band
Genuinely negotiable, the main lever
Recruiter with hiring manager sign-off
Base salary above the band ceiling
Rarely, and only as a named exception
Reward team plus a senior approver
Level or title band
Occasionally, and usually only before the offer
The panel or a levelling forum, not the recruiter
Sign-on or joining bonus
Often the most movable cash
Hiring manager or recruiter within a limit
Equity or share grant size
Sometimes, and often in a separate pool from cash
Reward team, sometimes an executive
Annual bonus target percentage
Usually fixed by level
Policy, effectively not negotiable
Bonus payout mechanics
Not negotiable, but ask how it works
Policy
Pension or retirement contribution
Almost never above the scheme rate
Policy
Holiday allowance
Occasionally at senior levels
Policy, sometimes with an exception route
Start date
Very often negotiable
Hiring manager
Remote or hybrid arrangement
Sometimes, and worth getting written down
Hiring manager, within a policy
Notice or probation length
Rarely on paper, occasionally in effect
Legal or HR policy
Equipment or home-office budget
Frequently, and cheaply
Hiring manager
A formal review at six months
Often, if you ask for it in writing
Hiring manager
Relocation or visa support
Fixed by an existing policy or absent
Policy
Two patterns in that table are worth extracting. Anything described as a policy is worth asking one clarifying question about and then dropping, because arguing with a policy costs goodwill and yields nothing. Anything within the hiring manager's gift is worth asking for, because they want you and the cost to them is small.
Levelling, and why it beats negotiating
If you take one thing from this page, take this. Moving up a level typically changes your compensation more than any counter inside a band will, and it changes the base from which everything afterwards is calculated. A successful negotiation moves you within a band. A successful levelling argument moves the band.
It also changes non-monetary things that compound: the scope you are given, the projects you are considered for, who you report to, and how long it takes to be considered for the next step. A person hired one level below where they should have been spends their first two years demonstrating what was already true, and re-levelling internally is slower and more political than being hired at the right level in the first place.
When level is decided, which is earlier than you think
Level is usually settled during the loop, in the debrief, before anybody talks to you about money. The panel forms a view from the interviews about the scope you can handle, whether you lead or contribute, how much ambiguity you can absorb, and whether you have operated at the level before. That view gets written down, and the offer is generated from it.
This has a consequence most candidates miss entirely: the highest-leverage moment in a compensation negotiation happens during your technical and behavioural interviews, before compensation is mentioned. If you want to be levelled senior, the evidence that gets you there is the scope of the examples you choose, whether your stories are about doing the work or about setting the direction, and whether you talk about outcomes across teams or tasks within one. A candidate who tells competent stories about their own tickets will be levelled at the level of somebody who does their own tickets, and no amount of firmness on a phone call afterwards will undo it.
Finding out what level you are being offered
Ask, plainly, at the verbal offer. Many candidates never do, and then negotiate a number without knowing which container it came from.
Before we talk about the number, can you tell me what level this offer is at, and how that maps to your internal framework? It would also help to know where in the band this offer sits, roughly, and what the range for that level looks like. I would rather understand the structure than guess at it.
Some organisations publish their levelling framework and will send it to you. Some will describe it. Some will refuse, which is itself information. The question is entirely normal and no reasonable recruiter is annoyed by it.
Arguing for a different level
The argument that works is against their criteria, not against your years. "I have eight years of experience" is not an argument, because the framework does not pay for years, it pays for scope. The argument that works picks the criteria for the level you want and points at evidence that you have already met them.
Thanks for this, and I am genuinely keen. I want to raise the level rather than the number, because I think that is the more useful conversation.
Your description of the senior level talks about owning a system end to end, setting technical direction for a team, and being the person other engineers escalate to. For the last two years that has been my job rather than my aspiration — I owned the payments integration from design through to the on-call rota, I set the migration approach that three teams then followed, and I was the named escalation point for that domain.
I am not asking you to take that on trust. If it would help I can walk whoever makes the levelling call through any of it, or answer written questions. But I think the level is the thing worth getting right, because it shapes the scope I am given more than the salary does.
Three things make that message work. It opens with commitment. It argues against their own stated criteria rather than against the outcome. And it offers evidence rather than asserting a conclusion, which gives the person on the other end something to take to the panel.
Be realistic about the odds. Re-levelling after an offer succeeds a minority of the time, because it requires someone to reverse a documented panel decision. It succeeds most often when the panel was split, when you have new evidence they did not see, or when the hiring manager already thought the level was wrong. The scenario is worked through in detail at offered a level below what you interviewed for.
When the level is fixed, ask for the route instead
If the answer is no, the fallback is worth more than another attempt at the same argument. Ask for the path, and ask for it in writing.
Understood, and I am not going to keep pushing on it. What I would like instead is to understand the route. What does the promotion process look like from this level, what evidence would I need to have, and when is the earliest cycle I could be considered in? If my manager and I could agree the specific things I would need to demonstrate, and write those down before I start, that would give me most of what I wanted from the level conversation.
That request is cheap for them, useful to you, and it converts a refusal into a commitment. It also tells you something: a manager who will not write down what good looks like at the next level may not have a clear idea, which is worth knowing before you join.
Reading an offer properly
An offer is not a number. It is a bundle of promises with very different degrees of certainty attached, and the headline figure a recruiter quotes has usually been assembled by adding them together as though they were the same kind of thing. Your first job is to take it apart.
The discipline that makes everything else work: separate what you will definitely receive from what you might receive. Guaranteed cash on one side. Everything conditional on performance, on the company's performance, on you still being there, or on a future event, on the other. Then decide whether the guaranteed side alone is a package you could live on and would accept. If it is not, the rest is a lottery ticket attached to a job you cannot afford.
Base salary
Base is the only component that is unconditional, and it is therefore the component worth the most per unit. It is what your mortgage lender looks at, what your bonus is calculated as a percentage of, what your pension contribution is a percentage of, and what your next employer will benchmark against. A given amount moved from base into a one-off payment is worth less to you and costs the employer less, which is exactly why they will offer to do it.
Check what the base actually is against everything you were told verbally. Check whether it is quoted before or after any deductions the employer makes at source, and whether any allowances quoted alongside it are contractual or discretionary. An allowance described in the letter as discretionary can be withdrawn, and often is when policy changes.
Check when the first review happens. An offer with a review three months after joining is worth more than the same offer with a review fourteen months away, and joining just after an annual review cycle can mean waiting the better part of two years for a rise. Ask when the cycle runs and whether new joiners are eligible in their first one — a surprising number of schemes prorate or exclude recent starters, and nobody volunteers this.
Bonus, and how it is really paid
The bonus percentage in an offer letter is a target, not a promise, and the mechanics between the target and the money are where the value is decided. Ask four things.
What determines the payout. Most schemes multiply a company component by an individual component. If the company component was zero in a recent year, the individual component was irrelevant. Ask what the scheme has paid out over the last few years — not as a promise, but as a history. A recruiter who will not answer is telling you something.
Whether it is discretionary or contractual. A discretionary bonus can be nil in a year the company chooses, for any reason or none. Most are discretionary. That is normal and not a reason to walk, but it means the bonus should be discounted heavily when you compare offers.
When it is paid and what you must be to receive it. Almost all schemes require you to be employed and not under notice on the payment date. If you join in one month and the payment date is another, you may be ineligible for the first period, or eligible only pro rata. If you later resign shortly before a payment date you lose the whole thing, which is a real consideration when timing a departure.
Whether it is prorated for your first year. It usually is. A target percentage applied to four months of service is a quarter of what the headline implies, and candidates routinely budget for the full figure.
Equity, and what makes it worth something
Equity divides sharply into two cases, and conflating them is the single most expensive mistake in offer evaluation.
Equity in a publicly traded company is close to cash with a delay and a price risk. The shares have a public price, you can generally sell them once they vest and once any trading window permits, and the main uncertainties are the share price, the tax treatment, and whether you stay long enough to vest. It is reasonable to value it, discounted for the price risk and for the probability you leave early. It is not reasonable to value it at the price on the day of the offer and treat that as salary.
Equity in a private company is a claim on a future event that may never happen. There is no market, so there is no price, and the number the recruiter quotes was produced by multiplying your share count by a price from the last funding round — a price set in a negotiation between the company and an investor, for a class of share that usually has protections yours does not. It may be worth many times that. It is more often worth nothing.
The honest position, and the one worth internalising: take a private-company offer only if the cash component alone is one you would accept. Treat the equity as an upside you would be delighted by and have not counted on. Everything that makes this concrete is at an offer that is mostly private company equity, and the questions to ask are these.
Question to ask
What the answer tells you
What instrument am I being granted, and what are the mechanics of turning it into shares?
Whether you face a decision and a cost later, or whether shares simply arrive
How many shares, and out of how many total on a fully diluted basis?
Your actual percentage; a share count alone is meaningless without the denominator
What price were the shares last valued at, and by whom?
Whether the quoted value comes from a real transaction or an internal estimate
What class of share is it, and what rights does that class have on a sale?
Whether investors are paid out before you, and by how much
How much money has been raised, and on what preference terms?
The size of the stack that has to be cleared before ordinary shares are worth anything
What is the vesting schedule and is there a cliff?
How long you must stay before any of it is real
If I leave after vesting, how long do I have to act, and what does it cost me?
Whether leaving forfeits vested value, which is common and rarely volunteered
Has there been any liquidity for employees, and on what terms?
Whether the company treats employee equity as real or theoretical
What happens to my grant if the company is acquired?
Whether acceleration exists, and whether a sale below a threshold pays ordinary holders nothing
Do I get a refresh grant, and on what basis?
Whether your equity decays after the initial grant vests
The preference question deserves a sentence of explanation because it is the one candidates most often skip and it is the one that most often decides whether the equity is worth anything. Investors typically buy a class of share that is paid out before ordinary shares on a sale. That means a sale price that looks like a success in a press release can clear the investors' entitlement and leave very little for ordinary shareholders. A large amount raised at a high valuation is not straightforwardly good news for an employee holding ordinary shares. Tax treatment of equity varies enormously by jurisdiction and by instrument, and can turn a paper gain into a real bill before any money exists; for anything material this is worth paying a professional for rather than reasoning about from a blog post.
Vesting shapes and cliffs
Vesting is how the promise turns into ownership, and its shape changes the value more than the headline size does.
Loading diagram…
The node worth staring at is the one after vesting. In a private company, vested is not the same as valuable, and a great many people have fully vested grants that have never been worth anything and may never be.
A cliff means nothing vests until a set point, typically the first anniversary, at which a chunk vests at once and the remainder vests periodically afterwards. Leave a day before the cliff and you have nothing. This is a retention device and it is normal, but it means the true cost of leaving early is higher than it looks and it should factor into how confident you are about the role.
The shape after the cliff varies and matters. An even schedule across four years is the common default. A back-loaded schedule, where a small proportion vests in the early years and a large proportion in the last, is a much stronger lock-in and is worth noticing, because the average annual value implied by the headline is not what you will receive in years one and two. Ask for the year-by-year breakdown rather than the total.
Ask what happens on leaving, in both directions. If you resign, unvested equity is normally forfeited, and vested equity may need to be acted on within a short window at a cost you have to fund yourself, which is how people end up walking away from grants they earned. If you are made redundant, terms are sometimes better, and it is worth knowing whether they are.
Sign-on bonuses and clawbacks
A sign-on bonus is the component companies find easiest to grant, for the reason above: it costs one year's budget and does not compound. If your counter is above what they can do on base, this is very often where the money reappears, and taking it is usually the right call.
Read the clawback. Almost all sign-on payments are repayable if you leave within a set period, and the terms vary in ways that matter. Ask whether the repayment is the gross amount or the net amount you received, because being asked to repay a gross figure out of which tax was already deducted is a genuinely unpleasant surprise. Ask whether it tapers over the period or is repayable in full until the last day. Ask whether it is triggered only by resignation or also by dismissal, and whether redundancy is excluded. Ask when it is paid — a sign-on paid after six months is a retention payment, not a joining payment.
The same questions apply to any relocation support, any training or certification the employer funds, and occasionally to visa costs. These are the clauses people sign without reading and discover eighteen months later. What to check in the offer letter before signing works through the document line by line.
Notice, probation and the terms nobody reads
Probation is a period during which the notice either side must give is shorter, sometimes a week or less, and during which some benefits may not apply. Check its length, check what notice applies during it, and check whether anything you care about — a bonus, a pension contribution, income protection — is suspended until you pass it. If you are leaving a secure position, a long probation with a one-week notice period is a real risk and worth weighing.
Notice period after probation cuts both ways. A long notice protects you if things go wrong and traps you if a better opportunity appears. Three months is common at senior levels and is a meaningful constraint on your next move.
Garden leave and restrictive covenants deserve a careful read. Non-compete and non-solicit clauses vary enormously in enforceability by jurisdiction, and a clause being unenforceable does not stop it being used to intimidate. If the clause is broad enough to affect where you could work next, that is a matter for a professional to look at rather than a matter for your own judgement, and the time to look is before signing.
What the letter refers to but does not contain. Offer letters routinely incorporate handbooks and policy documents by reference. Those documents are part of what you are agreeing to and you are entitled to read them. Ask for the ones that matter: the bonus scheme rules, the remote-working policy, the expenses policy if you will be travelling.
Pension, benefits and the things worth real money
Benefits are usually presented as a list and are usually skipped, which is a mistake, because two of them are worth substantial money and the rest are worth very little.
Retirement or pension contribution is the big one. An employer contribution is deferred pay and a difference of a few percentage points is worth more than most of the base difference candidates argue about. Find out the employer contribution, whether it is conditional on you contributing a matching amount, whether there is a cap, and whether there is a waiting period. Compare it properly against your current arrangement rather than treating it as an equivalent line item on both sides.
Health and insurance cover varies from nominal to genuinely valuable depending on where you are and what you would otherwise pay. Check whether cover extends to family, whether there is an excess, and whether anything meaningful is excluded. In some places this is worth a large fraction of a salary difference and in others it is worth almost nothing; work out which case you are in rather than assuming.
Income protection and life cover are cheap for the employer and occasionally very valuable to you. Worth checking, not worth negotiating.
Everything else — the perks, the allowances, the discounts — is generally worth less than it sounds and should not move a decision. Treat a long benefits list as a presentation choice rather than as value.
Then normalise. Put both offers into one comparable annual figure, with the guaranteed components at full value, the conditional components discounted by how likely you think they are, and the private-company equity at nothing or near it. Then look at what the numbers say and let the difference be as big or small as it truly is. Comparing two offers when one pays more works through the arithmetic and, more importantly, through what to do when the normalised numbers are close.
What is negotiable and what is policy
The practical test for whether something is worth pushing on: does granting it require a person to make a decision, or does it require a document to change? A hiring manager deciding to fund a better laptop is a decision. A pension contribution above the scheme rate is a document. Push on decisions. Ask one clarifying question about documents and then let them go, because arguing with a policy signals that you do not understand the organisation you are joining, which is a bad first impression to make for something you will not get.
The negotiation itself
Everything up to here is preparation. This section is the conversation.
Who names a number first
The conventional advice is never to name first, and it is roughly right but for the wrong reason. The reason is not that first movers lose. It is that they currently know less than you do, because they have a band and a levelling decision and a budget, and you have a guess.
So try once to get them to name it, without being cagey about it.
Before I put a number on it, could you tell me the range for this level? I would rather work inside your structure than guess at it, and it would make this a much shorter conversation.
If they give you a range, you have won the exchange, and your ask should sit in the upper part of it with a reason attached.
If they push back — and many recruiters are trained to push back exactly once — do not fight about it. A second refusal makes you look difficult over a point of technique. Give a number, and give it well.
Give a specific figure rather than a range, because a range is heard as its bottom. If you say you are looking for somewhere between two figures, the offer will come at the lower one and you will have negotiated against yourself. Say the figure you want, attach the reason, and stop talking.
Based on what I have seen for this level and this scope in the market, and on where my current package sits, I am targeting [figure] on base. I have some flexibility in how the total is put together — if base is constrained I am open to talking about how the rest is shaped — but that is the number I have in mind for the whole package.
Note what that does. It names one number. It justifies it against the market and the role rather than against your expenses. And it signals flexibility on structure without signalling flexibility on the total, which is exactly the flexibility you want to advertise.
Anchoring honestly
Anchoring works, which is why it is worth doing, and it is worth doing without lying. An anchor is credible in proportion to the reasoning attached to it, and an anchor with no reasoning is just a large number that invites a small counter.
The honest version: pick the top of the range you can justify with evidence, and be able to say where the evidence came from. Published salary data for the role and level in your market, what comparable roles have advertised, what you know of the band, what another offer says. Then name that figure as your target rather than as your minimum.
The dishonest versions all have the same failure mode. Inflating your current salary is a lie that background checks and payslip requests sometimes catch, and being caught ends the process and occasionally the offer. Inventing a competing offer is discussed below and is a worse bet than it looks. Naming a number you would not accept and then accepting far less teaches the recruiter that your numbers do not mean anything, which matters for the rest of the conversation and for every pay conversation you have there afterwards.
There is a floor worth deciding privately before any of this: the number below which you would decline. You do not say it out loud, ever, because saying it makes it the offer. But knowing it stops you from negotiating past your own limits in either direction.
A justified ask against a demand
The difference is not tone, though tone matters. It is structure. A justified ask contains a number, a reason external to you, and an open door. A demand contains a number and a consequence.
I would sign today at [figure].
That is a demand, even said pleasantly, because it presents a condition. It sometimes works. When it fails it fails hard, because you have made a statement you now have to either honour or retreat from, and both are bad.
Everything about the role is right and I want to do it. The one thing I would ask you to look at is the base. At [figure] I would be entirely comfortable, and that number comes from what I have seen for this scope at this level rather than from anything about my situation. Is there room to get there, or is there a different way to close the gap?
That is an ask. It contains the same number. It attaches the reason to the market rather than to your rent, which matters because your costs are not their problem and citing them invites a conversation about whether you can afford to live where you live. And it ends with a question that offers them two ways to say yes, one of which is the sign-on bonus they were probably going to suggest anyway.
Never justify a number by what you need. A candidate who says they need a certain figure because of a mortgage has moved the conversation to their circumstances, where the employer has no obligation and considerable discomfort. Justify by the market, by the scope of the role, by another offer if you have one, or by the band itself.
Ask for the package, not the base
The commonest tactical error is negotiating one component at a time, in sequence. You ask for more base, they improve it slightly, you accept, and then you ask about the sign-on, and then you ask about the start date. Each ask individually is reasonable. Together they are exhausting, and the third one gets a colder answer than the first, because the recruiter now believes there is no end to this.
Make one consolidated ask. Everything you want, in one message, with a clear indication of what matters most.
Thank you, and I want to say clearly that I want this job — the team and the problem are exactly what I was looking for.
There are three things I would ask you to look at, and they are in priority order. The base is the main one; at [figure] I would be very comfortable, and I have based that on what I have seen for this scope at this level. Second, if the base is constrained, a joining payment that closes part of the gap would work for me. Third, and much smaller, I would like to start four weeks later than the date in the letter, because I want to hand over properly rather than leave a mess behind me.
None of those are conditions. I would like to work with you either way, and I would rather put everything on the table at once than come back to you three times.
That message does the work. It is unambiguous about wanting the job, which removes the recruiter's main fear. It ranks the asks so they can grant the cheap ones if the expensive one fails. It hands them the sign-on solution rather than making them propose it. It puts a nice reason on the start date. And the last line explicitly buys you credit for not being a serial negotiator.
How many rounds is normal
One substantive round, sometimes two. That is the honest answer and it is lower than most advice implies.
Round one is your consolidated counter. Round two, if it happens, is a response to a partial grant — they moved on base but not as far as you asked, and you either accept or make one small closing ask. There is rarely a legitimate round three, and going there converts you from a candidate negotiating into a candidate who is difficult, a reputation that follows you into the job and is remembered by the manager who granted the last concession.
If they meet your number in full, accept. Do not go back for more because it was easy. A recruiter who meets a number and then receives a second ask learns that agreeing with you produces more work, and that lesson gets applied for the rest of your time there.
Countering, in the actual words
The situation
What to say
What it concedes
Standard counter with a target in mind
One message, ranked asks, market justification, explicit enthusiasm
Nothing; it names a figure without making it a condition
No competing offer and you know it
Justify entirely on scope and market, offer flexibility on shape
Signals you have no alternative, which they may already assume
You need time to decide
Name a specific date and a reason for it
Reveals that you are considering something else
They say it is the top of the band
Ask what would need to be true for a higher number
Accepts the band, moves to level or to non-cash
They say best and final
Accept, or make one small non-cash ask, or decline cleanly
Ends the money conversation either way
You are ready to accept
Confirm the whole package in writing, warmly, with no new asks
Nothing; closes it
The scripts below are written to be sent roughly as written, by email rather than on a call where possible. Email is better for this because it gives the recruiter something they can forward to an approver, and because it stops you from talking past the moment you have made your point.
The standard counter
Hello [name],
Thank you for the offer, and please pass on my thanks to [manager] and the panel. I want to be clear from the start that I want this role — the scope is what I have been looking for and the conversations have been the best part of my search.
There is one thing I would like you to look at. The base is a bit below where I expected to land for this level. At [figure] I would be entirely comfortable, and that number comes from what I have seen for equivalent scope in this market rather than from anything about my own situation.
If base is constrained by the band, I am open to how the package gets there — a joining payment, a larger equity grant, or an earlier review would all be reasonable ways to close it from my side.
Is there room to move, and what would be the most workable route for you?
Happy to talk it through on a call if that is easier.
Countering with no competing offer
You do not have to have an alternative to negotiate, and you should not pretend to have one. What you have instead is a reason, and a reason is enough for a first counter.
Thanks for this, and I am pleased — I want to do this job.
I want to be straightforward with you: I am not running another process to the same stage, so this is not me weighing you against somebody else. It is that I have looked at what this scope pays at this level and the offer sits below where I expected, and I would rather say so now than accept and be quietly unhappy about it in six months.
[Figure] on base is what I had in mind. If that is not possible, tell me and I will not keep pushing — but if there is room, I would like to use it.
The reason to name the absence of an alternative rather than to leave it ambiguous is that the ambiguity does not help you. A recruiter assumes you are interviewing elsewhere in any case, and an unstated implication is not leverage. What the honesty buys is credibility for everything else in the message, which is the thing that gets a counter approved when nothing external is forcing it. It also removes any temptation to answer are you interviewing elsewhere in a way you cannot support later.
Asking for time
Thank you — this is genuinely good news and I am glad it worked out.
Could I have until [specific day] to come back to you? I want to read the full documentation properly, including the bonus scheme and the equity terms, and I have one conversation I promised to finish before I make a decision. That is [number] working days, and I will give you an answer on that day either way rather than going quiet.
If that date is a problem for any reason, tell me what would work and I will fit around it.
Name a specific day. "A few days" invites a chase and a misunderstanding; a date is a commitment they can plan around. Give a reason that is true, and "I have another process to finish" is an acceptable reason to give — it is a fact, not a threat, provided you do not attach a consequence to it. And promise to answer on the day, then do.
Asking what would need to be true
This is the most useful question in the whole negotiation and almost nobody asks it. Use it when you have been told a number cannot move.
That is helpful, thank you, and I will take the band as given rather than arguing about it.
Can I ask a different question. What would need to be true for the offer to be at a higher number? Is it a different level, is it a different scope of role, or is it something about my background that the panel wanted more evidence on? I am not trying to reopen anything — I would genuinely like to understand where the constraint is, because if it is something I can address I would rather know now than discover it in a year.
Three outcomes, all useful. Sometimes the answer is "a different level, and the panel was split", which is an invitation. Sometimes it is "the band, and that is that", which lets you stop cleanly and move to the non-monetary asks with your goodwill intact. And sometimes it is a specific gap the panel saw, which is worth more than the money, because you now know what you will be working against from your first week.
Handling a best and final
"This is our best and final offer" is sometimes literally true and sometimes a closing technique. You usually cannot tell the difference, and the good news is that you do not need to, because the correct response is the same either way.
Do not test it with another cash ask. If it is true, you look like you were not listening. If it is not true, you have started a second round on a point they have publicly closed, which forces them to either lose face or hold.
Understood, and thank you for being direct about it — I would rather know than keep going round.
I am happy to accept at that number. Two small things that are not about the money, and neither is a condition of my signing: could we make the start date [date] rather than [date], and could you confirm the two remote days we discussed in the offer letter rather than leaving them as an understanding? Both of those matter more to me than the last part of the gap did.
Send the paperwork over and I will get it back to you quickly.
That reply accepts, which removes the tension, and then asks for two things that cost nothing and are almost always granted because the recruiter is relieved. It is also the right shape if you were going to accept anyway: you have converted a closed conversation into two small wins.
If the number is genuinely below what you can accept, say so cleanly rather than continuing to negotiate. "I understand, and I appreciate you being straight with me. At that number I am going to have to say no, which I am sorry about, because I would have enjoyed the work. If anything changes, I would be glad to talk again." That sometimes produces a further movement and sometimes does not, and either way it is the honest version.
Accepting
Thank you — I am delighted, and I accept.
Just so we both have the same record, my understanding of the package is: base of [figure], [bonus] target bonus subject to the scheme, a joining payment of [figure] repayable on a tapering basis if I leave within [period], [equity detail], starting on [date], with two days a week remote as we agreed. Could you confirm that is right and send the contract over?
I have not yet given notice and I will do that once the contract is signed and the background check has cleared. Please let me know what you need from me for that.
Really looking forward to it — thanks for how straightforward this has been.
The restatement is the important part and it is the step people skip because they are pleased and want to stop. Verbal agreements decay, recruiters leave, and the only version of the arrangement that survives a change of personnel is the written one. Setting out your understanding in a message they reply to costs nothing and has saved a great many people a difficult conversation in month three.
When to stop
Stop when the marginal money is smaller than the cost of asking for it. That cost is not zero and it is not only reputational: every extra round delays your start, extends the period in which something can go wrong, and consumes the goodwill of the person who will decide what you work on.
Concretely, stop when they have met you, when they have moved substantially towards you, when they have said best and final, or when you have made two asks and received two answers. Stop immediately if the tone changes — a recruiter who becomes formal, or who starts referring to "the process" rather than to what they can do, is a recruiter who has stopped advocating for you.
And stop if you notice that you are negotiating to win rather than to be paid properly. That is a real failure mode and it costs people offers. The question to ask yourself is whether you would accept the current offer if the alternative were no job. If the answer is yes, the remaining conversation is about margin, and margin is not worth much risk.
Competing offers
A second genuine offer is the strongest position in this whole subject, and it is also the one most commonly mishandled.
Using one without bluffing
The rules are simple. Say that you have another offer. Do not say the number unless it is higher and you are prepared to be held to it. Do not ask them to beat it. Say clearly which one you would prefer, if that is true, because it is the sentence that makes the whole thing work.
I want to be open with you about where I am. I have a second offer, and the package is meaningfully ahead of yours. I would rather work with you — the problem is more interesting and I got on with the team — and I would like to find a way to make that the easy decision rather than the expensive one.
I am not asking you to match it, and I know the structures are different. But if there is room to close some of the gap, that would let me say yes to you without feeling I had ignored the difference. Where the number is now, it is a harder call than I want it to be.
The reason this works better than a matching demand is that it gives the hiring manager something to argue with. "Close some of the gap so we can have him" is a case a manager can take to a reward team. "Match this or he goes" is an ultimatum, and reward teams refuse ultimatums as a matter of policy because granting one guarantees the next one.
Saying you prefer them is the load-bearing sentence and it must be true. It converts the conversation from a bidding war, which they will lose or refuse, into a request to remove an obstacle, which they can grant.
Why bluffing is a bad bet
Inventing an offer, or inflating one you have, is common advice and a poor wager. Consider the ways it goes wrong.
They call it. "Which company? What is the number? What is the deadline?" is a normal set of questions and there is no comfortable answer if the offer does not exist. Recruiters ask these routinely, not because they suspect you, but because they need the information to build a case internally.
They ask for it in writing. Some organisations will not process an exception without seeing the competing offer letter. This is less common than it used to be but it happens, and the request is unanswerable.
They believe you and act on it. Now you have an offer contingent on a fiction, and if the fiction ever surfaces — through a mutual contact, through a recruiter who works both accounts, through an offhand comment in your first month — you have a credibility problem inside the company rather than outside it.
They believe you and let you go. This is the outcome people forget. A company that cannot match your invented number may conclude that they cannot afford you and withdraw gracefully, and you have lost a real offer to a fake one.
And the smaller cost: the industry is not as big as it feels. Recruiters move, hiring managers know each other, and specialist markets are small.
The honest alternative is available and almost as good. "I am at final stage with one other company" is true if it is true, and it is sufficient to create urgency without creating a number you cannot support.
Aligning timelines
The reason most people do not have competing offers is not that they were not competitive. It is that the processes finished three weeks apart. Aligning them is a scheduling problem and it is largely solvable, but only if you start early.
Loading diagram…
The branch that matters is the honest one on the right. Sometimes the second process cannot move, and the correct answer is then to decide on the offer in front of you rather than to gamble a real offer against a possible one. Candidates lose good offers waiting for a process that was never going to finish in time.
What to do in practice. Start the processes you care about within the same fortnight where you can, because a three-week head start is very hard to close later. When the first offer arrives, ask for a decision date rather than accepting the default. Then tell the other process immediately, without drama and without a number.
I wanted to let you know where I am rather than leave you guessing. I have received an offer elsewhere and I have until [date] to respond. I am still very interested in this role and I would much rather have both to compare properly. Is there any realistic way to reach a decision by then? If there is not, I completely understand, and I would rather know than hold you to something that does not fit your process.
That message is effective for a specific reason: it is not a threat, it contains a real date, and it gives them an easy way to say no. Processes accelerate for it more often than candidates expect, because a hiring manager who wants you would much rather compress two interviews into one week than lose you to a scheduling accident.
If they cannot accelerate, believe them and decide. And if you decide against them, tell them, because the person you handled well is the person who calls you next year.
Exploding offers and deadline pressure
An exploding offer is one with a deadline short enough that you cannot make a considered decision or finish another process. Twenty-four hours, forty-eight hours, "we need an answer by Friday" delivered on Thursday afternoon.
Sometimes the deadline is real. A contractor is leaving, a project starts, a headcount expires at the end of a quarter, a second candidate is waiting for an answer. Where it is real, the recruiter can say what drives it, in one sentence, without hesitating.
Sometimes the deadline is a technique to stop you from comparing. Where it is a technique, the reason is vague, it changes when questioned, or it is delivered with an appeal to your enthusiasm — "we are looking for someone who is excited enough to know straight away".
So ask, and treat the answer as data about the company as much as about the deadline.
Thank you, and I am glad. Can I ask what is driving the date? I ask because I want to give you a proper answer rather than a rushed one, and if there is something specific behind it — a start date, a headcount cycle — I will work to it.
Then ask for a specific extension, with a reason and a commitment attached.
Could I have until [day] instead? That is [number] working days. I want to read the bonus scheme and the equity documentation properly before I sign something, and I would rather do that than come back with questions after I have accepted. I will give you a firm answer on that day, and if it is a no I will tell you straight away rather than letting it run.
Reasonable employers grant this almost always. An employer who refuses a short, specific, reasoned extension to sign an employment contract has told you how they will handle every reasonable request you make once you work there, and that is worth more than the offer is. The full treatment of this scenario, including what to do when the extension is refused, is at offer with a twenty-four hour deadline.
Two things not to do. Do not accept with the private intention of reneging if something better arrives; it is a small industry, it does real damage, and it is worse than a clean refusal. And do not use another company's exploding deadline to pressure a process that is going well, because the pressure lands as a threat and it is remembered.
Negotiating the things that are not money
When the number will not move — and often it genuinely will not — there is a second list, and much of it is easier to get because it does not touch a band, does not create a documented anomaly and does not require an approver above the hiring manager. It is also frequently worth more to your actual life than the last increment of base.
The ask
How often it moves
How to frame it
A later start date
Very often
A proper handover to your current employer, which they respect
An earlier start date
Often, if they are under pressure
Offer it as a concession in exchange for something else
Specific remote or hybrid days in writing
Often, within an existing policy
As confirming what you discussed rather than as a new ask
Equipment or home-office budget
Often
A specific item and what it is for, not an allowance
A formal review at six months
Often
Framed as a checkpoint, with what would be assessed
Title, where it does not change the level
Sometimes
On external credibility or client-facing need, not status
A training or conference budget
Sometimes
Named and costed, tied to something the role needs
Written scope, or a named first project
Sometimes
As mutual clarity rather than as a demand for guarantees
Extra holiday
Occasionally, senior levels
Usually only where a policy exception route exists
Notice or probation length
Rarely
Only worth raising if the terms are unusual
Three of these deserve more than a table row.
A start date is the easiest thing on the list and the most under-used. Framing it around handing over properly to your current employer is a framing that works because it demonstrates the behaviour they would want from you when you leave them. It also buys you a gap between jobs if you want one, which is worth considerably more than a small salary difference and is almost impossible to get later. Where visa or right-to-work timing is involved, the start date interacts with an authorisation process that has its own clock, and that is covered in the visa and right to work guide rather than here. The related timing questions — notice, relocation, remote — are worked through at when to raise notice period, relocation and remote expectations.
A written commitment about scope is the most valuable non-monetary ask and the least often made. Not a guarantee, which no manager can honestly give, but a written statement of what the role is understood to be and what you would be doing first. The reason to want it is not legal. It is that the single commonest source of early regret is a role that turns out to be different from the one described, and writing the description down is the cheapest available test of whether the manager has a clear idea. A manager who cannot write a paragraph about what you will own in your first quarter has told you something important for free.
One last thing, and it is not a negotiation point. Could you and [manager] write me a short paragraph on what the first six months looks like — the system or area I would own, the first project, and what a good six months would look like from your side? Partly so I can hit the ground running, and partly because I would rather we found out now if we are picturing different jobs.
A review at six months is the standard fallback when the base is at the band ceiling or the level is fixed. It is worth having only if it is specific: a date, who conducts it, what evidence would be considered, and what the outcome could be. A vague promise that "we will revisit it" is worth nothing and is granted freely for precisely that reason.
If the number is fixed, would you be willing to put a formal review at six months in the offer letter? What I would find useful is a date, agreement that a re-levelling or an out-of-cycle adjustment is on the table at that point, and a short note from [manager] on what they would want to see. I am not asking for a promise about the outcome — just that the conversation happens on a date rather than when someone remembers.
When not to negotiate
Not every offer should be countered, and advice that says otherwise is not being honest about the cases where pushing costs more than it gains.
When the offer is already at or above what you asked for. If you named a figure and they met or beat it, countering tells them your numbers are not real. Accept and be pleased.
When you have been explicitly told the number is fixed and you have already tested it once. A second attempt on the same point is not persistence, it is not listening.
In a formal graduate, apprentice or cohort programme. These pay a set rate to every entrant by design, and the recruiter genuinely cannot change it. Asking once is harmless if phrased as a question; pressing is a bad start with people who will be your peers on the same rate.
In a small organisation where the whole cash position is visible. In an early-stage company, the founder may genuinely be at the limit of what the company can pay, and pushing hard on cash can read as a misunderstanding of the situation you are joining. Push on equity structure and on scope instead.
Where the process is regulated or the pay scale is public. Some sectors publish their scales and place you on a point by a defined rule. The negotiation there is about which point on the scale you enter at and whether prior experience counts, which is a factual argument rather than a bargaining one.
Where the relationship is already strained. If the process has been difficult, if you pushed back on something earlier and it landed badly, if a stakeholder was lukewarm, the offer may be more fragile than it looks. This is a judgement call and not a rule, but it is worth making consciously rather than negotiating on autopilot.
Where the difference is genuinely small. If the gap is a rounding error against the total, weigh it against the cost of the exchange, which includes several days of delay and a slightly worse start.
And the honest general caveat: a well-phrased single counter very rarely costs you an offer, but "very rarely" is not "never". Some hiring managers read any negotiation as difficulty, some organisations have a culture where asking is unusual, and a small number of offers have been withdrawn over a counter that was aggressive in tone. You cannot get the benefit without accepting a small residual risk. What you can do is make the risk as small as possible: one ask, warm, justified externally, explicitly not a condition, with clear enthusiasm for the job attached.
Rescinded offers, and what raises the risk
Offers are withdrawn, and while it is uncommon, it is worth knowing what raises the probability, because most of it is within your control.
Conditions that do not clear. Most offers are conditional on references, right-to-work documentation, and often a background or credit check depending on the sector. The most common cause of a withdrawn offer is not a bad reference, it is a discrepancy — dates that do not match, a qualification described differently on the CV than on the certificate, an employment gap presented as something it was not. Any of this is survivable if disclosed and fatal if discovered, which is the entire argument of anything a background check will turn up.
A hiring freeze or a budget change. Entirely outside your influence and the reason for the advice at the end of this page about not resigning too early. Headcount can disappear between a verbal offer and a contract, particularly near the end of a financial period.
A negotiation that read as bad faith. Accepting and then reopening, a competing offer that turns out not to exist, an ultimatum delivered and then not honoured. These are rare and they are the ones candidates cause.
Something that surfaced late. A public post, a reference volunteering something unexpected, a mutual contact offering an unsolicited opinion. Little to do here other than knowing your references and what they will say.
The role changed. Reorganisations happen mid-hire, and the role you accepted occasionally becomes a different role. If you are told the scope has changed materially before you start, treat that as a new offer to be evaluated rather than as an administrative detail.
The practical defences are all timing. Get everything in writing. Do not resign until the contract is signed and every condition is cleared. Do not make an irreversible financial commitment — a lease, a move, a resignation from a stable position — on the strength of a verbal offer, however warm the call was.
After acceptance, and before you resign
The gap between accepting and starting is where the avoidable disasters happen, and all of them are avoided by the same three habits.
Get it in writing, all of it. Every verbal commitment — the remote days, the review date, the start date, the first project, the equipment budget, the exclusion of redundancy from the sign-on clawback — either appears in the offer letter or in an email the recruiter has replied to affirmatively. The person who promised it may not be there in a year. If a recruiter is reluctant to put something in writing that they were happy to say out loud, that is the most useful signal you will get in the entire process, and the correct response is a mild, curious question about why.
Read the whole document before signing, including what it refers to. Reconcile it line by line against what you were told, and query every difference rather than assuming it is a typographical error. Some of them are. Some of them are the actual terms, and the verbal version was optimistic. Ask for the policy documents the letter incorporates. If anything material is unclear — restrictive covenants, intellectual property assignment, a clawback with unusual terms — that is worth a professional opinion, and the cost of one is trivial against the cost of being wrong about it.
Do not resign until it is signed and clear. This is the single most important sentence in this section. Not until the contract is signed by both parties, references have come back, the background check has cleared, and any right-to-work or visa condition is satisfied. Resigning on a verbal offer is the most expensive mistake available at this stage, and the fact that it usually works out does not make it a good bet, because the case where it does not is catastrophic and the cost of waiting a week is nothing.
When conditions do take time — and background checks in some sectors take weeks — say so to your new employer and agree a start date that accounts for it, rather than trying to compress your notice period to fit.
Resigning, and the counter-offer
Two things happen after you resign, and the second is the one to plan for.
Resign in writing, briefly, without a list of grievances. The letter is a formality that goes in a file and it is not the place for the reasons. Tell your manager in a conversation first, before the letter and before anyone else hears it. Offer a proper handover and then actually do one, because your leaving behaviour is what your former colleagues will remember and reference conversations are less formal than people think.
Then, in a meaningful proportion of cases, a counter-offer arrives.
How to think about a counter-offer
Start with why it exists. Your employer has just discovered that replacing you costs a recruitment fee, a vacancy of some months, a ramp-up period, and the risk of a worse hire. Against that, a rise is cheap. The counter-offer is a rational response to a cost they have just been shown, and that is not the same thing as a reassessment of your value.
The question that cuts through it: what does the counter-offer say about what they thought of you last month? If the money was available, it was available before you resigned, and it was not offered. If your scope could be expanded, it could have been expanded when you asked. A counter-offer is evidence about what your employer will do when threatened, which is a different and less useful fact than what they will do when not.
Then check whether it addresses your actual reason for leaving. Most people leave for a reason that money is a proxy for rather than a cause of: the work, the manager, the trajectory, the sense of being stuck. A counter-offer almost always addresses the money because money is the thing an organisation can move in a week. If your reason was the work, and the counter is a rise, then in six months you will be doing the same work for slightly more, which is where most people who accept a counter-offer end up. If it is a genuine change of role, with a different manager or a defined new scope, that is a different proposition and deserves to be considered on its merits.
There are also two costs that are real and are usually understated. You have shown you were prepared to leave, and in some organisations that changes how you are regarded — quietly, in the conversations about who gets the next interesting thing. And you have burnt the other offer, which will not be there in six months when you conclude that nothing changed.
The version of this question you will also face in the interview room is if your employer matched the offer, would you stay, and the reason interviewers ask it is precisely this: they are pricing the risk that they will make you an offer, you will resign, and your employer will keep you. A candidate whose reason for leaving is not about money is a candidate whose acceptance will hold. That is the same reason why this company and why are you leaving gets asked as carefully as it does.
If you decide to decline the counter-offer, say so quickly and warmly, and do not use it as leverage anywhere.
I really appreciate this, and I want you to know it means something that you have put it together. I am still going to go, and the reason is not the money — it is the [scope or direction], and that is not something a rise changes. I would rather be straight with you than let this run on. What I would like to do is make the handover as good as I can, and I would like to leave on terms where we would both be glad to work together again.
And if you decide to accept it, do so with your eyes open about what has to change beyond the number, and get that in writing too, for exactly the reasons in the section above.
What interviewers ask
Compensation conversations contain a small number of recurring questions, and almost none of them are requests for information. Each is protecting something or testing something, and knowing which changes the answer.
The question
What it is really doing
What a good answer contains
What are your salary expectations?
Getting you to anchor first, and screening you out of the band
A researched figure or a deflection to their range, tied to market
What are you currently earning?
Anchoring on your last employer's valuation of you
A redirection to what you are looking for now
Are you interviewing elsewhere?
Pricing urgency and the risk of losing you
A truthful stage, no number, no threat
If we matched, would you sign today?
Checking whether the number is the only obstacle
Enthusiasm plus honesty about what else is open
That is the top of the band. How flexible are you?
Testing whether your number was real
A move to structure or level, not a collapse
Would you take the offer at the lower level?
Testing whether title matters more than the work
A clear position, plus a request for the route
What would make you sign today?
Trying to close, and to surface every remaining ask
Your full remaining list, once, in priority order
If your employer counter-offers, what happens?
Pricing the risk of a wasted offer
A non-money reason for leaving, stated plainly
How soon could you start?
Planning, and occasionally testing your notice discipline
A real date that respects your notice period
The cross-cutting signal is consistency. Every number and every date you give is written down somewhere, and the recruiter has the notes from your screening call in front of them during the offer conversation. A figure that moves, a competing offer that changes shape, a notice period that shrinks when it becomes convenient — each of these individually is small and together they cost you the benefit of the doubt at exactly the moment you need it.
The second cross-cutting signal is whether you can hold a position without becoming unpleasant. This is not politeness for its own sake. The hiring manager is deciding what it will be like to disagree with you about a deadline in eight months, and they have exactly one piece of evidence: how you behaved when you disagreed with them about money.
Questions
The questions below are phrased the way recruiters and hiring managers put them. Each answer names what the other side is assessing or protecting, because in a compensation conversation the question is very often not about the information it appears to request.
What are your salary expectations for this role?
Try once to turn it around, then answer properly if pressed. "Before I put a figure on it — what is the range you have budgeted for this level? I would rather work inside your structure than guess. If it helps, I am not going to be difficult about it, I just do not want to name a number that turns out to be miles off in either direction." If they will not, give one figure with a market justification and stop.
What they are protecting is the band and their own time; a candidate whose expectations are far above the range is a candidate they want to disqualify early rather than after four rounds. What they are assessing is whether your number is researched or improvised, and improvisation is audible. The failure is a wide range, which is heard as its lower bound, and the second failure is anchoring on your current salary, which prices you at your last employer's valuation rather than at this role's. The full treatment is at what are your salary expectations.
What are you currently earning?
Answer the question you would rather have been asked, without refusing in a way that creates friction. "I would rather talk about what I am looking for than what I am on, because my current package reflects a role I have outgrown rather than the scope of this one. For this level I am targeting [figure], and I have based that on the market rather than on a percentage uplift."
The protection here is a cheap anchor. If they know your current number, every subsequent offer is calibrated against it plus a small increment, regardless of what the role is worth. In many places there are rules limiting whether this can be asked or used, and the shape of them varies enough that you should not rely on any particular one applying to you; the practical answer works whether or not a rule protects you. What is being assessed is whether you can decline something without sounding either evasive or combative, and the redirect achieves both because it offers information rather than withholding it.
Are you interviewing anywhere else, and how far along are you?
Tell the truth about the stage, without a number and without a threat. "Yes, two other processes. One is at final stage and I would expect to hear in the next couple of weeks; the other is early. I am not trying to create a race — I am telling you so we can plan the timing sensibly, because I would rather have everything to compare properly than make a decision under pressure."
What they are pricing is urgency and risk: how fast they need to move, and how likely they are to lose you. What they are assessing is whether you use a competing process as a weapon, because a candidate who does that in a negotiation will do it in a pay review. The most common mistake is exaggerating, and the reason it is a mistake is that the details get asked for. This question is worked through at are you interviewing elsewhere.
If we could get to your number, would you sign today?
Do not say yes unless it is literally true. "If the base got to [figure] and everything else stayed as it is, then yes, subject to reading the contract and the bonus scheme properly — which I will do quickly, not over a fortnight. I would not want to say yes and then come back with something else, so let me be complete: the base is the only open item for me."
They are trying to establish that the number is the last obstacle, which is what they need in order to justify the exception internally. A recruiter who spends political capital getting your figure approved and is then met with a new ask has been embarrassed, and they remember it. What is being assessed is whether you have put all your cards on the table. An honest "yes, with these two small things also outstanding" is a much better answer than a yes you then qualify.
That is the top of the band for this level. How flexible are you?
Accept the band and change the subject to structure. "Thank you for telling me that, and I will take it at face value. Two questions then. Is there flexibility on the joining payment or the equity, which I understand often sits outside the base band? And separately — what would need to be true for this to be at a different level? I am not reopening it, I just want to understand where the constraint is."
They are protecting the band's integrity and internal equity, and they are also checking whether your number was a real position or an opening bid you will abandon at the first resistance. What is assessed is whether you can distinguish between a constraint and a refusal. A candidate who repeats the same number louder has shown they do not understand how pay is set; a candidate who moves cleanly to the components outside the band has shown they do.
We can put the difference into a joining bonus rather than base. Does that work?
Usually take it, but be clear about the difference and check the terms. "That works, thank you. Two things I want to be clear on before I say yes. It is a one-year fix rather than a permanent one, so I would want to know when the first base review is and whether I am eligible in the next cycle. And what is the recovery period on the joining payment, is it gross or net, and does it taper? If it is repayable in full for two years I would weigh it differently."
What they are protecting is the recurring cost and the band, both of which a one-off payment leaves untouched. What is being assessed is whether you understand the difference between one-off and recurring money, and whether you read terms. Accepting a sign-on without asking about the clawback is the commonest unforced error in this whole subject, and the follow-up about the review cycle is what stops a one-year fix from becoming a permanent shortfall.
The written offer is below the number you gave us earlier. What do you want to do?
Point at the gap without accusation and make one specific ask. "It is, and I would like to understand it rather than assume anything. When we spoke I said [figure] and you said that was workable, so I want to check whether something changed in the levelling or whether this is where the band landed. Either way, my position is the same: at [figure] I am in, and if base cannot get there I am open to how the package makes it up."
They may be protecting a level decision that came out lower than expected during the loop, or they may be testing whether your number was firm. What is being assessed is whether you notice, whether you raise it calmly, and whether one clean counter follows rather than a sequence of asks. The detailed version is at the written offer is below the number you gave.
We are offering the level below the one you interviewed for. How do you feel about that?
Ask what they saw before you argue. "Can I ask what the gap was? Not to relitigate it, but because it is useful to me either way. If the panel wanted more evidence of something specific, I would like to know what." Then, if you have the evidence, argue against the criteria rather than against the outcome. If the level is fixed, ask for the route to the next one in writing.
What they are protecting is a documented panel decision, which is expensive to reverse. What is being assessed is whether you can receive a downgrade without either collapsing into acceptance or becoming aggrieved, and whether your response is about the work or about the title. A candidate who asks what would need to be demonstrated and gets it written down has recovered most of the value even when the level does not move. Fully worked at offered a level below what you interviewed for.
Why do you think you are worth more than what we have offered?
Answer with the market and the scope, never with your circumstances. "It is less about me and more about the role. Everything I have seen for a role with this scope at this level sits above where the offer landed, and I would rather say that now than accept and quietly wonder about it. If your data says something different I would genuinely like to see it, because you have better information than I do."
They are protecting the band and testing your justification. What is being assessed is the basis of your number. A candidate who cites their expenses has moved to ground where the employer has no obligation and considerable awkwardness. A candidate who cites the market has made a claim that can be checked, which is a much stronger position, and inviting them to correct you with their own benchmark data is a confident move that occasionally produces a useful answer.
This is our best and final offer.
Accept it or decline it, but do not test it with another cash ask. "Understood, and thank you for saying so directly rather than letting me keep pushing. I am happy at that number. Two small things that are not about money and are not conditions — could we move the start date to [date], and could the two remote days go in the letter rather than staying as an understanding?"
They are protecting a decision that has probably already been approved by someone else, and closing the conversation. What is being assessed is whether you can accept a closed door gracefully, which is a direct preview of how you will handle a decision that goes against you at work. Pushing again on cash after a best and final is the one move that reliably damages the relationship, and it rarely produces money.
How quickly can you make a decision?
Give a specific date and a reason. "I would like until [day], which is [number] working days. I want to read the bonus scheme and the equity documentation properly rather than skim them, and I have one conversation to finish. I will come back to you on that day either way — I will not go quiet on you."
They are protecting a timeline that may be real, and they are also gauging how likely they are to lose you. What is being assessed is whether you can commit to a date and hold it. A vague "as soon as I can" invites a chase and reads as either disorganisation or as manoeuvring. Naming the day, and then delivering on it, is worth a surprising amount of goodwill going into the job.
The offer expires at the end of tomorrow.
Ask what is behind it, then ask for a specific extension. "Can I ask what is driving that date? If there is a start date or a headcount cycle behind it I will work to it. If there is flexibility, what I would like is until [day] — I want to read the documentation properly before I sign an employment contract, and I will give you a firm answer on that day."
They may be protecting a real constraint, or they may be preventing you from comparing. What is being assessed is whether you can be pressured into signing something you have not read, and some organisations quite deliberately want to know. The most important thing here is that the answer is information about them: a refusal to grant a short, reasoned extension to read a contract is a preview of the culture. See offer with a twenty-four hour deadline.
You have another offer. What is the number?
Give it only if it is genuinely higher and you are willing to be held to it. Otherwise decline without evasiveness. "I would rather not put their number in the middle of this, partly because the structures are not comparable and partly because I do not think it is quite fair to them. What I can tell you is that the total is meaningfully ahead of yours, and that I would rather be here. If you can close part of the gap this becomes an easy decision for me."
They need the number to build an internal case, which is a legitimate reason to ask. But handing it over turns the conversation into a matching exercise that reward teams often refuse on principle. What is being assessed is whether the offer is real, and the tell is not the number but the confidence and consistency of everything around it. If you invented it, this is the question that exposes you.
Would you accept if we matched your other offer exactly?
Answer honestly, which usually means saying what else matters. "If the packages were the same, yes, I would come here, and I would tell them today. It is not only the money — the problem here is more interesting to me and I got on with the team. But I would not pretend the difference is nothing, which is why I am asking."
They are checking whether money is genuinely the deciding factor, because if it is not, matching is wasted money. What is being assessed is whether your stated preference is real. A candidate who says the money is irrelevant is not believed. A candidate who says they prefer this role and would like the gap closed enough to make it easy has given them a reason to spend, and has said something that is checkable against their behaviour.
Most of this package is equity in a private company. How do you think about that?
Separate the cash from the option and be honest that the equity is contingent. "I look at the cash on its own first and ask whether that alone is something I would accept, because the equity might be worth nothing and I would rather know now whether I could live with that outcome. Then I want to understand the structure rather than the headline number: what instrument it is, the share count against the fully diluted total, the preference stack ahead of ordinary shares, the vesting shape, and what happens if I leave. If the answers are good I treat it as real upside. I just do not count it as pay."
They are protecting a recruiting tool that only works if candidates value it, and some will be uncomfortable with the questions. What is being assessed is commercial literacy and whether you can be sold a large number. A candidate who accepts the quoted value uncritically has told them something about their judgement generally. The complete list of questions is at an offer that is mostly private company equity.
The equity is valued at the last funding round's price. Does that not make it real?
Say what the price is and is not, without being dismissive. "It is a real price for a real transaction, but it is the price an investor paid for a different class of share with protections mine would not have, and it is a price at one moment. What decides whether my shares are worth anything is the preference stack ahead of them and what the company eventually sells for. I am not saying it is worthless — I would just want to understand the structure before I treat that figure as part of my salary."
They are protecting the headline total they used to make the offer competitive. What is being assessed is whether you understand what you are being granted. This is one of the few questions where a well-informed answer materially improves how you are seen, because it demonstrates the same commercial judgement they will need from you in the role.
How would you feel about a smaller base with a larger bonus opportunity?
Price the certainty difference and ask about the history. "It depends entirely on the scheme. What has it paid over the last few years, is it discretionary or contractual, and what determines the company component? If it has paid consistently and the mechanics are clear, I am open to it. If it is discretionary and has been nil in a recent year, then I am being asked to take a base cut for a possibility, and I would want the base to reflect that."
They are protecting a cost structure that shifts risk from the company to you, which is a legitimate thing to want and a legitimate thing to price. What is being assessed is whether you distinguish guaranteed from contingent money. Accepting on the headline is the failure. Asking for the payout history is the move, and the answer — or the refusal to answer — tells you most of what you need.
Is base the only thing you want to move, or is there more?
Give the complete list, once, ranked. "Base is the main one. Beyond that: a start date four weeks later than the letter, and the two remote days written into the offer rather than left as an understanding. That is everything — I would rather give you the full list now than come back to you three times."
They are trying to establish the total cost of closing you so they can seek one approval rather than three. What is being assessed is whether you negotiate in one round or in a sequence, and the sequence is what turns a cooperative recruiter into a defensive one. Holding something back to deploy later is a tactic that works once and damages the relationship permanently.
What would it take for you to sign right now?
Answer literally and completely, or say plainly that you are not ready. "Two things, and then I would sign. The base at [figure], and the remote arrangement in writing. If both of those happen I will sign the same day." Or, if you are genuinely not ready: "Honestly, nothing today — I have another conversation to finish and I would rather not commit before it. What I can give you is a firm answer by [day]."
They are closing. What is being assessed is whether your answer is a real specification or a moving target. If you name your terms and they are met, you have to sign, and a candidate who then produces a new condition has done real damage. The alternative answer — that no number closes it today because something else is open — is entirely acceptable and much better than a specification you will not honour.
If we made this offer, would your current employer be able to keep you?
Answer no, and give the non-money reason as the evidence. "No. I have thought about it, and the reason I am leaving is not the money — it is that the platform work I want to do does not exist in my current role and is not going to. A rise would not change that, and I would rather be straight with you than have you find out in three weeks."
They are pricing the risk of making an offer that gets used as leverage elsewhere, which happens often enough that many recruiters ask it routinely. What is being assessed is whether your reason for leaving is one that a counter-offer can address. A hedged answer costs you goodwill and occasionally costs you the offer, because a candidate who would stay for money is a candidate whose acceptance is not safe. Detailed at if your employer matched the offer, would you stay.
You have changed jobs quite often. What makes this different, and should we be worried about the notice period?
Address the pattern first and the logistics second. "The two short stays both had a specific cause — one was a reorganisation three months after I joined and one was a role that turned out to be different from the description. What I do differently now is exactly what I have been doing with you: asking what the first six months looks like, in writing, so we both find out early if we are picturing different jobs. On notice, mine is [period] and I intend to serve it properly."
They are protecting against a costly early departure, and the offer stage is where that risk gets priced. What is being assessed is self-awareness and whether the pattern has an explanation that does not blame every previous employer. The related question is worked at frequent job changes, why will this one be different, and the adjacent one about how you left at was leaving your last role your decision.
When could you start, and is there any flexibility on that?
Give a real date built from your notice period, and offer the flexibility you actually have. "My notice is [period], so realistically [date]. If you need me earlier I can ask about being released a week or two early, but I would rather not promise it before I have asked, and I do want to hand over properly. If a later date suited you better, [date] also works for me."
They are planning, and occasionally testing whether you would leave your current employer badly, which is a preview of how you would leave them. What is being assessed is whether your dates are real and whether you respect an obligation. Promising a start date that requires you to abandon your notice period is a bad signal delivered in the guise of enthusiasm. The related timing discussion is at when to raise notice period, relocation and remote expectations.
We cannot change the base. What else would make a difference to you?
Take the invitation seriously, because it is a real one, and ask for specific things. "That is helpful, thank you. Three things, none of them expensive. A start date of [date], the two remote days written into the letter, and a formal review at six months with a date on it and a note from [manager] on what would be assessed. If those are possible I am very happy."
What they are protecting is the band; what they are offering is a route to close you without touching it, and a recruiter who asks this question wants to give you something. What is being assessed is whether you can shift from a closed door to an open one. A candidate who has nothing to ask for here has wasted the invitation, and a candidate who tries to reintroduce the base by another name has misread it.
Can you put in writing that you will accept if we get to that number?
Be careful, and be honest about what you can commit to. "I can tell you clearly that if the base is [figure] and the rest of the package is as we have discussed, I will accept, and I will do it the same day the contract arrives. I would rather say that to you plainly than sign something conditional in advance, because until I have read the contract I do not know what is in it. If it helps, I am happy to put that sentence in an email."
They are seeking certainty before spending an approval, which is reasonable. What is being assessed is whether your word is worth something and whether you understand what you are being asked to commit to. Committing to a package you have not read is a bad idea. Committing to a number, in writing, with the contract review reserved, is a fair position and it is usually enough for them.
There is a clause here you have queried. It is standard and we cannot change it. What now?
Decide whether it is material, and say which. "Understood. Most of what I asked about was clarification and I am content. The one that matters to me is the joining payment recovery — as drafted it is repayable in full for two years, gross, including if I were made redundant. I am not asking to remove it, but could redundancy be excluded and could it taper? If neither is possible I would want to know that and I will think about it."
They are protecting a template that has been through legal review, and "standard" is frequently accurate. What is being assessed is whether you can distinguish a term you dislike from a term you will not accept, and whether you raise it as a question rather than as an accusation. A candidate who queries fifteen clauses has signalled how they will handle every process document they meet. A candidate who queries one, specifically, with a proposed narrowing rather than a deletion, is behaving like a professional.
We would like to move the start date forward by a month. Can you do that?
Treat it as an exchange rather than a favour. "I can look at it. My notice is [period] and I would need to ask to be released early, which my manager may or may not agree to — I do not want to promise something that is not mine to give. If I can make it work, would you be able to [do the thing you wanted]? I am not trying to trade, but if I am asking my current employer for a favour I would rather it bought something."
They have a real constraint, usually a project or a departure. What is being assessed is whether you will over-promise under enthusiasm, and whether you will treat your current employer's interests as disposable. This is also one of the cleanest opportunities in the whole negotiation, because a company asking you for something is a company in a mood to give something.
You went quiet for a week after we made the offer. What happened?
Explain without over-apologising, and reconfirm your position. "I said I would come back by [date] and I did, but I should have sent you a holding note in between — that is on me. What I was doing was waiting on the other process, which I told you about, and it took longer than they said it would. My position has not changed: I want to do this job."
They are reading how you handle a commitment and whether the silence means you are being shopped around. What is being assessed is communication discipline, which is a direct proxy for how you will behave when a deliverable slips. The recovery is a short acknowledgement and no elaborate excuse. Ghosting during an offer window is the single behaviour most likely to make a recruiter quietly stop advocating for you, and it is entirely avoidable with a two-line email.
Do you have any questions for us before you decide?
Ask the ones whose answers change your decision, not the ones that make you look interested. What the first six months looks like and what you would own. When the pay review cycle runs and whether you are eligible in the next one. What the bonus scheme has paid recently. What happened to the last person in the role. How the team has changed in the last year. And, if you want to know one thing about the manager, what they would want you to have achieved by the end of your first quarter.
What is being assessed is what you care about, and there is a real difference between a candidate whose questions are about the work and a candidate whose questions are entirely about the package. Both are legitimate; the mix tells them something. What you are assessing is whether the answers are specific. A manager who cannot describe your first quarter concretely may not know what the role is yet, and the offer stage is the last cheap moment to find that out — the same point made in where do you see yourself in five years, where the useful answer depends on the role having a shape somebody can describe.
You did not get the level you wanted and you took the job anyway. Any regrets?
If this comes up later, in a review or a subsequent interview, answer it without resentment. "I took it because the work was right and I got the route to the next level written down before I started, which is what I actually wanted from the level conversation. It has been about what I expected. If it had not been, I would have raised it at the six-month review that we agreed, which is the reason I asked for the review to have a date on it."
What is being assessed is whether a decision that did not go your way has become a grievance. A candidate still litigating their levelling eighteen months later has told an interviewer something about how they hold disappointment, which is the same thing being examined in the interview rejection and self-diagnosis guide. The strong version treats the written route as the thing you negotiated for, because it was.