Build the total by listing every component with a consistent rule: base at face value, formula bonuses at target, discretionary bonuses at zero, equity divided across the vesting years, and benefits at their real cost to you. Then subtract what the job costs — commute, relocation, any benefit you lose. The single figure is a comparison tool rather than a decision, and its main value is exposing components you would otherwise never have priced.
Build the number with consistent rules
The whole exercise depends on treating both offers the same way, so decide the rules first and apply them without exception. Inconsistent rules produce a comparison that flatters whichever offer you already prefer.
Base salary goes in at face value, because it is the one component that is genuinely certain. Everything below it in the list carries some condition attached. A guaranteed bonus goes in at face value too. A bonus with a defined formula goes in at target rather than at maximum, since target is what the plan expects to pay in an ordinary year of trading. Maximum figures describe a year that nobody should plan around.
A discretionary bonus goes in at zero. That feels harsh and it is the only defensible treatment, because a payment the employer may choose not to make is not something you can plan around. If it arrives it is a good surprise, and if you counted it and it does not arrive you made a decision on money that never existed.
Equity goes in divided across the vesting years rather than at its headline total. A grant described as $80,000 vesting over four years is $20,000 a year, and comparing the headline against another offer’s annual salary is the single most common error in this whole exercise. Headline totals and annual figures are different units entirely.
Then subtract what the job costs you
An offer has costs attached and they belong in the same arithmetic. The commute is the largest for most people — not just fuel or fares, but the vehicle, the parking, and the hours themselves. Those hours are the largest cost and the one nobody ever prices if you want to price them properly. An hour a day each way is roughly six working weeks a year.
Relocation costs go in, and since 2026 employer-paid relocation is taxable wages for civilian employees, so the decisive question is whether any relocation package is grossed up. An unadjusted package is worth substantially less than its face value suggests. Ask directly whether the relocation figure quoted is grossed up.
Then any benefit you lose. Leaving before a vesting date, forfeiting an accrued leave balance, or moving from a plan with a strong retirement match to a weaker one are all real costs that never appear anywhere in an offer letter. You have to remember to go looking for them yourself.
The lines that resist a single number
Some things genuinely do not convert and pretending otherwise produces a false precision. Health coverage is the clearest case: two plans with identical premiums can differ enormously in deductible, network and out-of-pocket maximum, and which is better depends on what happens to you rather than on any figure you could compute in advance. Price the plan against the year you expect to have.
Paid leave converts arithmetically — a day is a day’s salary — and it does not convert in terms of what it is worth to you, which varies more between people than almost anything else in a package. Somebody with young children values it differently from somebody without.
Equity in a private company resists it most of all. A grant has a face value based on a valuation set for preferred shares with protections your common shares do not carry, and in many outcomes common holders receive substantially less than the headline arithmetic implies. Ask about the preference stack before treating that number as real.
Why one number is still worth producing
Given all that, it is fair to ask why bother. The answer is that the exercise forces you to price components you would otherwise ignore entirely, and that is precisely where the surprises live. Half the value here is the components it forces you to look up.
Most people compare base salaries and glance at everything else. Running the full calculation regularly reveals that the lower-base offer is ahead once the retirement match, the health premium and the leave are counted — and equally often reveals that a large headline equity number is worth far less annually than it first appeared. Both discoveries change decisions, and neither shows up in a base comparison.
The number also gives you something to negotiate against. Knowing that one offer is $6,000 behind on a consistent basis lets you ask for a specific figure rather than gesture at a feeling that something is short somewhere. Specific figures get answered and vague impressions get deflected.
A worked comparison
Take two offers. The first is $95,000 base, a ten percent target bonus, a four percent retirement match, and a health premium costing you $220 a month. The second is $88,000 base, a discretionary bonus, a six percent match, a premium costing you $60 a month, and five extra days of leave.
The first comes to $95,000 plus $9,500 target bonus plus $3,800 match, less $2,640 in premiums, giving $105,660. The second is $88,000 plus zero for the discretionary bonus plus $5,280 match, less $720 in premiums, giving $92,560 — plus five days of leave worth about $1,692 at that salary, for $94,252.
So the first offer is ahead by roughly $11,400 rather than the $7,000 the base salaries suggested. The bonus treatment is doing most of that work, which is exactly why the rule about discretionary bonuses matters so much. One judgment call moved that comparison by four thousand dollars.
What the number should not decide alone
It should not decide anything on its own, and treating it as a verdict is how people talk themselves into jobs they were uneasy about. The number is one input among several, and the others are frequently larger over five years. Treat the total as one column in a much wider decision.
The work itself, the person you would report to, what you would learn, and whether the role leads anywhere all matter more over five years than a difference of a few thousand dollars. A job that teaches you something scarce is worth more than the gap between two offers, because scarcity is the thing that moves pay most durably across a career. A scarce skill outlasts any single offer you compare it against.
The check that reveals the ceiling
One question turns a static comparison into a forward-looking one. Ask what the band is for the level you are being hired into and where in it your offer sits today. Those two answers describe your next three years rather than your first one.
An offer at the bottom of a wide band has room above it without any level change. The same figure at the top of a narrow band means your next increase requires a promotion, which is a different proposition entirely. Two identical offers can have completely different futures for that reason, and the only way to find out is to ask.
Write it down
Put the components in a list with your rule beside each one, and keep it. Two things follow from that.
The first is that you can hand the reasoning to somebody else — a partner, a friend — and have them check it. That catches the errors that come from wanting one particular answer. The second is that when the next offer arrives in three years, you have the method rather than having to reconstruct it.
And write down the assumptions rather than only the totals. A comparison that depended on valuing a discretionary bonus at zero should say so, because a year later you will not remember which choices produced the number and whether they still hold.
Common questions
How should I treat each component?
Base and guaranteed bonuses at face value, formula bonuses at target, discretionary bonuses at zero, equity divided across the vesting years, and benefits at their real cost to you.
Why value a discretionary bonus at zero?
Because a payment the employer may choose not to make is not something you can plan around. If it arrives it is a good surprise; if you counted it, you decided on money that never existed.
What is the most common error?
Comparing an equity headline against an annual salary. A grant of $80,000 vesting over four years is $20,000 a year, and the two figures are not the same kind of thing.
What costs should I subtract?
Commute in full, relocation — which since 2026 is taxable wages for civilians, so ask whether it is grossed up — and any benefit you lose, including forfeited vesting or leave.
Can you show a worked comparison?
$95,000 with a 10% target bonus, 4% match and $220 monthly premiums comes to $105,660. $88,000 with a discretionary bonus, 6% match, $60 premiums and five extra leave days comes to $94,252.
What does that example show?
A gap of about $11,400 rather than the $7,000 the base salaries suggested. The treatment of the discretionary bonus is doing most of that work.
What resists a single number?
Health coverage, where identical premiums can hide very different deductibles and networks; paid leave, whose value varies by person; and private-company equity, priced off preferred shares you do not hold.
What question reveals the future?
Where your offer sits in the band for the level. The bottom of a wide band has room without a promotion; the top of a narrow one means the next increase requires one.