Every few months the question recedes. A holiday resets you, a new manager arrives, and the urgency drops for a few weeks. Then it comes back, more insistent than before. You keep looking up the same search terms, hoping different phrasing will change the answer. It never has.

Why symptom checklists fail

Most articles that try to answer this question ask you to rate a list of feelings. How you feel on Mondays. Whether you dread your inbox. How often you think about quitting. The list looks diagnostic, but it skips the question that matters: which conditions produced those feelings. A bad quarter, a hard manager, a year of bad sleep, or a career that does not fit you all produce the same set of symptoms. Your mood in January and your mood in April can produce opposite conclusions about the same job, and neither reading holds up on its own.

What you need is evidence that does not flip when the conditions do. Even then, the evidence is a hypothesis about your situation, not a verdict. Anyone claiming otherwise, including you at 11pm on a Tuesday, is overreaching.

What conditions can and cannot tell you

One place to look is your own attention. What you used to follow about your field on your own time, whether that was reading, listening, watching, or talking to other people in it. Locate when that stopped.

If you never gave your field voluntary attention in the first place, skip this test. The two signals below can run without it.

The test is whether the stopping tracks with conditions. Managers change, projects end, and pressure eventually lets up. If your interest in the work recovered during any of those shifts, treat that as evidence for a phase. Better conditions changed the response, which means conditions were doing the work.

The harder case is when conditions have shifted meaningfully and the interest has stayed gone. Treat that as evidence for a fit problem. Not proof. Conditions are never perfectly clean, and burnout takes longer to recover from than a good quarter or two. If you have been running depleted for years, the recovery window has to be longer than a season for the reading to mean anything.

Separate the career from the job around it

Before concluding the career itself is wrong, check what else could be producing the reading. The wrong employer with a decent role in your field will feel like the whole field is wrong. So will the wrong level, the wrong workload, the wrong team.

You do not have to run a full audit. Ask whether a version of this same career at a different employer, a different level, or a different workload would feel meaningfully different to you. If you can picture a version that would, the problem is probably the job configuration. If no version appeals to you regardless of the setting, the reading points at the career itself.

A second signal, and a third

Alongside interest recovery, two other signals are worth checking.

One: whether the doing itself still produces satisfaction. Set aside outcomes and recognition and look at the moment you are actually doing the work. Is anything alive in it? If a decent stretch of conditions gives you back the outcomes but the work still feels flat while you are doing it, that flatness is data.

Two: adjacent-role envy. When you hear about people doing slightly different versions of your work, do you feel drawn or relieved you are not them? Notice which adjacent versions pull you. When one specific version pulls hard, the mismatch may be narrower than the whole career. When every adjacent version repels you as much as your current one, the pull is somewhere outside your field entirely.

Neither signal answers on its own. The question is whether they point the same way as the first test.

Three readings of what you see

Reading one. Your interest recovered during a decent stretch, and the other signals held up in the version of the job you would want. Treat this as a conditions problem. The work is fine. Something in the surroundings is producing the reading.

Reading two. Your interest did not recover during a decent stretch, adjacent versions do not pull, and the doing itself has stayed flat. Treat this as a fit hypothesis, load-bearing enough to act on.

Reading three. You cannot find a decent stretch in the last twelve months to test against, or the signals contradict each other. Most self-run diagnostics arrive here, and it is the honest place to sit. You do not have data yet.

What to do with each

If it is a conditions problem, the next step is with the conditions. Change what you can change, and re-check the reading once you are somewhere you can breathe.

If it is a fit hypothesis, the temptation is to exit the field this quarter. That is a trap. What the diagnosis calls for is narrowing the direction to look before the search consumes another year, and holding more evidence before you act. Slower is the right speed here. If you've already made the change once and it turned out wrong, the second decision starts differently

If it is insufficient evidence, the next step is getting yourself into six to eight weeks of decent conditions and running this again. That is harder than it sounds. You are probably depleted, probably in a job configuration that does not hand you a clean window, and probably impatient with a question that has been circling for months. The alternative is answering the question from inside the conditions that produced it, which is how you have ended up here each time. The premature version of this question is what has kept you looping.

The search stops when the diagnosis holds. Whether it holds for the outcome you were expecting is a separate question.