Comparing job offers beyond salary

Salary dominates job comparisons for a boring reason: it is the only factor that arrives as a clean number. Everything else has to be estimated, so it quietly drops out of the comparison, and the offer with the larger headline wins by default rather than on merit.

Look at the table above. Offer A pays four thousand dollars more in base salary. Offer B pays about seven thousand four hundred more once the bonus, the employer retirement contribution, the health premium, and the learning budget are counted, and closer to eleven thousand once the extra ten days of leave are valued at a day rate. Nothing in that comparison is exotic. It is arithmetic on numbers both employers already published, and it reverses the answer that base salary alone would have produced.

There is a second cost to the salary-only habit. The factors it excludes are the ones that determine how long you stay. A role you leave in eleven months resets your earnings trajectory, costs you a search, and looks worse on your record than the pay difference was worth. Tenure is part of the financial calculation even when you evaluate it purely as money.

None of this argues that salary is unimportant. It argues that salary is the first line of the analysis rather than the whole of it.

Total compensation is every form of value the employer transfers to you in a year. Base salary, cash bonus at target rather than at maximum, commission at a realistic attainment rather than the number on the plan document, equity valued honestly, employer retirement contributions, the employer share of insurance, paid leave, stipends, and any reimbursement you would otherwise pay for yourself.

Two rules keep the arithmetic honest. Use target figures, not ceilings: a bonus described as "up to twenty percent" should enter your spreadsheet at whatever the team actually paid last year, and if nobody will state that number, treat the silence as information. And discount equity in a private company heavily. It may be worth a great deal, it may be worth nothing, and it cannot pay rent in the meantime, so it belongs in a separate line rather than added to cash as though the two are the same.

Also count the money that leaves. A role requiring five commuting days at an hour each way costs fuel or fares, plus roughly five hundred hours a year of your time. A relocation costs a moving bill and, often, a higher cost of living that erodes the raise that justified the move. Those are negative entries in the same table, not vague misgivings.

Once you have both columns, compare the priced totals rather than the salaries. That single change corrects most of the error in most offer comparisons.

Health coverage is usually the largest hidden line. Read three numbers rather than one: the premium deducted from each paycheck, the annual deductible, and the out-of-pocket maximum. A plan with a low premium and a high deductible looks cheap in a healthy year and costs thousands in a year with a surgery or a birth. Price the plan against the year you realistically expect, and if you have a chronic condition or a family, price the bad year too.

Retirement contributions are close to direct pay. An employer putting in six percent against another employer's three percent is a real difference of thousands per year, compounding for decades. Count the full amount, and read the vesting schedule, because a contribution you have to stay four years to keep is worth less than one you own immediately.

Paid leave has a defensible day rate: divide the base salary by about two hundred and sixty working days. Ten extra days on an eighty-eight thousand dollar salary is roughly three thousand four hundred dollars. Unlimited leave policies deserve scepticism rather than a high number, because what people take under them frequently falls below what an accrued policy would have granted. Value the policy at the amount employees actually take.

Everything else, meaning stipends, learning budgets, equipment allowances, and parental leave beyond the statutory minimum, is priced the same way: what would you have paid for it yourself.

Some of the largest factors resist pricing entirely, and the correct response is to rate them rather than to drop them. Your manager is the clearest example. The person you report to shapes your daily experience, your access to interesting work, and your odds of a promotion more than the company name does, and no dollar figure captures that.

The rest of the unpriced list is short and consistent: the people you would sit with, whether the work itself interests you, how decisions are made, whether the company is stable enough to still exist in three years, how much autonomy you have over your own day, and whether the culture treats boundaries as real or as an aspiration on a careers page.

Rate each of these one to five with evidence rather than atmosphere. Evidence means specifics you can ask about: how the last two promotions on this team happened, what an on-call week actually looks like, when someone last took two consecutive weeks off, why the previous person in this seat left. Vague answers to specific questions are themselves a rating.

Then weight them. A one-to-five rating on stability means nothing until you say how much stability matters to you relative to pay, which is exactly what a values assessment produces. Without weights, the unpriced factors either dominate the decision through mood or vanish from it through convenience.

Turn the question around, because the reversed version is far easier to answer honestly. Instead of asking whether the higher offer is worth the trade, assume you have already accepted the lower one and ask what you would pay to undo the difference.

Suppose Offer A pays eight thousand more but adds forty-five minutes each way. That is roughly seven and a half hours a week, or three hundred and fifty hours a year, so the premium is about twenty-three dollars per commuting hour before fuel. Compare that against what an hour of your evening is genuinely worth to you. Some people answer yes immediately; others realise they have been talking themselves into something.

The same reversal works for any trade-off. Would you pay eight thousand dollars a year for a manager you trust? For a role with a written promotion path? For a team where nobody messages on Sunday? These are uncomfortable questions precisely because they force a real number onto something you would rather keep vague, and that is what makes them useful.

One caution: the trade only holds while the pay difference is discretionary. If the higher offer is what covers your rent or your childcare, it stops being a values question and becomes an arithmetic one. Name that honestly rather than dressing a financial necessity up as a preference.

Sometimes the salary-first answer is the right one, and pretending otherwise is its own kind of dishonesty. If one offer does not cover your actual monthly costs, no amount of mission or autonomy repairs that, and the comparison is over before the scorecard starts.

Salary also deserves extra weight when you are carrying debt with a real interest rate, when you are supporting other people, when you are rebuilding after a gap in work, or when the gap between the two offers is large enough to change your options rather than merely your comfort. A thirty percent difference is a different kind of decision from a four percent difference, and treating both as "just one factor" is a mistake in the other direction.

The honest framing is that pay is a weighted factor whose weight is personal and changes over a life. For someone with savings and no dependents, compensation may sit third. For someone six months out of work, it may sit first by a wide margin, and their scorecard should say so rather than quietly overriding it later.

What stays constant is the method. Price what can be priced, rate what cannot, set the weights before the offers arrive, and read the result as a reason rather than a verdict.

Build one table with your factors down the left, your weight for each in the second column, and a column per offer. Priced items enter as a rating derived from the dollar totals rather than as raw dollars, so that a seven thousand dollar gap does not silently outvote every other row on scale alone. Multiply, total, and read the columns.

Two habits keep the exercise from deceiving you. Set the weights before the offers are on the table, and leave unknowns as unknowns rather than scoring them as average. A cluster of unknowns on your heaviest factors is itself a signal.

A two-point gap between two offers is noise. A twenty-point gap is a signal. Anything in between means the deciding factor is probably something you have not written down yet, which is usually a person, a commute, or a piece of the work you already know you would resent.