There is a style of career advice that presents a decision as a modelling exercise. Score your current role out of fifty. Build three five-year salary scenarios. Assign probabilities to outcomes and compute an expected value. Conclude, with two decimal places, that the transition is worth $155,000.
I have a doctorate and I model things for a living, and I want to be direct: applied to a career decision, this is mostly numerology. Not because quantifying is wrong, but because every input is invented. You do not know the probability that a pivot works out. You do not know what you would have earned by staying. Multiplying two guesses does not produce knowledge; it produces a guess with a false air of authority, and — this is the actual harm — it lets you feel you have analysed something when you have only formatted it.
There is a real analysis to do here. It is smaller and less satisfying than a spreadsheet, and it is about survivability rather than optimisation.
The one calculation worth doing
Ask how long you can go without the income, and what happens at the end of that period.
This is the only number in the whole exercise that is knowable, because it is arithmetic on facts you already have: savings, monthly expenses, other household income, obligations that do not pause. It also happens to be the number that determines whether a bad outcome is a setback or a catastrophe, which is the distinction that actually matters.
Build the buffer before you need it. The conventional starting point is three to six months of expenses, and it is reasonable to sit at the higher end — or beyond it — when you are about to make your income less predictable on purpose. There is no magic threshold; the point is that a long runway converts a career change from a gamble into an experiment, and experiments can be repeated.
Everything else in the standard framework is a way of dressing up preferences as calculations.
What the numbers genuinely settle
A few financial questions are real and specific, and people routinely miss them while building elaborate models of things that cannot be known.
Unvested equity, and what you forfeit by leaving before a cliff or a vest date. This is a concrete, knowable sum, and it is frequently the largest single financial fact in the decision.
Notice periods, non-compete or non-solicit clauses, and any training-cost or bond clause you signed. These are contractual and checkable.
Benefits that do not transfer: health cover for a family, particularly where continuity or pre-existing-condition waiting periods are involved. Visa or work-authorisation status tied to the employer, which in some situations makes the timing of a move a legal question rather than a financial one.
Pension or retirement contributions, and anything with a vesting schedule.
These deserve an hour with the documents. That hour is worth more than any number of scenario tables, because the answers exist.
Where quantification does damage
Three specific failures are worth naming.
Opportunity cost as a hard number. Articles compute the earnings you forgo by leaving, projecting your current salary forward with annual raises and a promotion in year four. That path is not a fact you are giving up; it is a hypothesis about a company that could restructure, a manager who could leave, or an industry that could contract. Treating it as a certainty and the new path as a risk systematically biases every comparison toward staying.
Probability estimates you have no basis for. "30% chance the role is not a fit" is a number produced by a mood. Feed several of these into a weighted expected value and the output is determined by your priors, laundered through arithmetic.
Scoring your own satisfaction. Rating six aspects of your job out of five, weighting them, and reading off a verdict does not add information — you supplied every input, including the weights, and you can feel which way you want the answer to go. What the exercise does do is let you avoid saying the true sentence, which is usually something like "I am bored and I am afraid I have left it too late."
What actually lowers the risk
Risk in a career move comes from irreversibility and from uncertainty about fit. Both can be attacked directly, and neither responds to modelling.
Test before you commit. Almost every transition has a partial version: a project in the target area inside your current job, contract work, a secondment, a few months of evenings building the thing. This is the single highest-value move available, because it converts an unanswerable question — will I like this, am I any good at it — into an observation. A striking number of people discover at this stage that they wanted the idea of the new field rather than its daily reality, and finding that out for the cost of a side project rather than a career is an enormous win.
Make the first step smaller. The riskiest version of a pivot changes employer, function and industry simultaneously. Change one at a time and each step is legible to the next employer, which keeps you hireable throughout. Same function, new industry, is usually easy. New function, same employer, is usually easy. Both at once, cold, is where people stall.
Have somebody on the inside. The probability of a transition working is heavily influenced by whether anyone in the new place is invested in your success. A manager who has hired a career-changer before, a colleague who will tell you what you are getting wrong in the first month. This matters more than the size of the pay cut.
Preserve the return path, deliberately. Keep your technical skills warm for a year. Stay on good terms with the people you are leaving. Do not burn the bridge in the exit interview, however tempting. The option to go back is worth a great deal, and it decays quietly — which is a genuine argument for moving earlier rather than later, since the decay is on the calendar rather than in your control.
The decision you should write down
Not a matrix. Two or three conditions, decided in advance, that would make you go — and one or two that would make you stay — written before you are emotionally committed.
The value is not in the criteria being correct. It is that later, when you are exhausted by the search or dazzled by a single flattering offer, you have a record of what you thought when you were thinking clearly. If you then override it, at least you will know that you are overriding something rather than discovering a new preference that happens to justify what you already want.
Set a date to look at it again. Quarterly is about right. Career decisions made continuously become a background anxiety that degrades your performance in the job you currently have, which is its own risk and rarely counted.
The part the frameworks cannot hold
Most of the real inputs here are not financial. Whether you are bored in a way that has started to affect your health. Whether you are staying because the work matters to you or because leaving would require admitting something. Whether the field you are looking at is genuinely interesting or merely better paid and more discussed.
These are not soft considerations to be assigned a weighting — they are usually the whole decision, and the spreadsheet exists to avoid them.
My own move out of academia was not the product of analysis. There was a period of about a year in which I kept noticing that the parts of the work I was most engaged by were the parts closest to something being used by someone, and eventually the conclusion was obvious rather than calculated. What I did do properly was make sure I could survive being wrong. That turned out to be sufficient, and I suspect it usually is.
Tags
Taresh Sharan