How to Know If You Should Look for a New Job

Most job-search advice collapses two decisions into one question. That is the category mistake. The threshold for opening the market is lower than the threshold for leaving the market you are in. Confusing the two is why people either quit too early (opened the market, took the first offer, landed in a similar role at a different company) or stay too long (never opened the market because they were not ready to quit, and the two decisions blurred into one impossible one).

A useful reframe: "should I look for a new job" is a lower-cost, higher-information move. It produces data, pay-band comparisons, interview loops, alternative role fits, that the "should I leave" decision needs as input. Trying to make the leaving decision without ever opening the market is like pricing a house without checking Zillow.

The question resolves to two parts. What signals mean it is time to open the market, and which signals look like it but are not?

Looking for a new job and quitting your job are two different decisions

Both decisions get treated as the same one. They are structurally distinct.

Looking is exploratory. It costs about 15 hours over a month, and the current job situation does not change during the process. The output is information: what the market rate is for the current title, what alternative roles look like when someone else runs them, whether the frustrations of the current role are ambient (they would follow the worker to any similar role) or structural (they are specific to this employer, this manager, or this trajectory).

Quitting is a commitment with a runway cost and a re-entry cost. It requires cash reserves, a specific target, and a decision about what the next role must provide that the current one does not. That decision is much harder to make well without the information the looking phase produces.

The workable sequence is to look for a new job first, then decide about leaving second. The muddled sequence, deciding whether to quit before ever opening the market, is what produces the worst outcomes on both sides. People who quit without market data land in similar roles for income reasons. People who never look because they are not ready to quit stay in place for years past the point when they had a real alternative available.

Four green-light signals that mean you should at least open the market

A green-light here is a structural condition. Moods reset. Structural conditions do not.

Learning-curve flat for six months or more. The role that taught things three years ago has no remaining learning curve. Meetings are predictable. Problems are re-runs. Competence is high, engagement is low. This is the wrong-stage archetype from the I hate my job cornerstone. Once the curve has been flat six months, working harder inside the role does not restore engagement. Only structural change does. Opening the market is one option; internal expanded scope is another. Both start with information the current role cannot provide.

Manager change that changed the job. New manager, changed working relationship, and the new equilibrium is one the employee does not want to sustain. Sometimes it resolves with a transfer request. Sometimes it does not. The market is where the alternative-manager alternative becomes visible. An interview loop under a hypothetical new manager surfaces what "same work, better relationship" actually looks like.

Pay drift past 15% below market for the same title. Comp letters compound. A 12% under-market position at year one is a 20-25% under-market position at year three because raises stack on the same base. According to a Pew Research survey of workers who quit a job in 2021, 63% named low pay as a reason and 63% named no advancement opportunities. Pay drift is one of the few career problems that arithmetically worsens on its own. Opening the market produces the number that resolves whether the gap is real.

Strategic re-org that shrinks the role. The company reorganizes, the role's scope narrows, and the trajectory changes in a way that will not revert within 12 months. Waiting to see if the re-org "settles" usually costs a year. The market at least prices the alternative during the year.

Two of these together is a strong signal to run the full process. One alone is enough to open the market and see what shows up.

Three fake green-lights that look like reasons to look but aren't

Not every reason to be frustrated is a reason to interview. Three of the most common ones are structurally noise.

One bad week. A specific hard Tuesday, a difficult project handoff, one meeting that went sideways. The dread here is state-dependent, and it resolves within two to three weeks in almost every case. Opening the market on a bad-week impulse produces low-quality applications and a distorted read on the current role. Wait two weeks. If the frustration is still specific and named, it might be signal. If it dissolved on its own, it was mood.

A single peer got promoted or left. Comparisons with a specific peer are almost always misleading. The peer's promotion may have been political, may have been role-narrowing rather than a genuine level-up, may have been the last one for that team for a while. The peer's exit may have been for reasons unrelated to the current role. Neither event tells the market anything. Broader comparisons are the actual data: SHRM salary bands, LinkedIn's median pay for the title, direct recruiter conversations.

Burnout, projecting onto the job. According to the Gallup 2025 State of the Global Workplace report, global employee engagement fell to 20% in 2025, the lowest level since 2020. That is a systemic condition that colors how any current job reads. Running a job search from a burned-out state produces two problems. The read on current-job frustrations is unreliable. The interview process itself requires energy reserves the burned-out state does not have. The sequence that works is to address the burnout state first, then reassess whether the job is the cause or the surface where burnout showed. The WHO ICD-11 entry on burnout as an occupational phenomenon is the framing that makes this distinction concrete.

The four-week quiet-look: resume, two recruiter calls, one interview loop, then decide

The quiet-look converts speculation into data at low cost, without changing the current job situation. Roughly 15 hours over four weeks. The steps are ordered.

Week 1: resume refresh. Two to three hours. Not a total rewrite. Update the last two years of accomplishments, verify titles, add any new skills or credentials. The resume is the artifact everything else runs on; a stale one produces stale conversations.

Weeks 2-3: two recruiter calls. Different specializations. One in the same industry, one in an adjacent one. Recruiters have a live read on hiring appetite, comp bands, and which titles are moving. A 30-minute call each produces market data no internal source can. The conversation is exploratory. The recruiter is not the target; the information is.

Week 4: one interview loop. One company, one role, all the way through to an offer stage if it goes that far. The process teaches more than the outcome does. Interview loops surface what the current role has taught, what the market values in the current title, and what "a different version of this job" concretely looks like when someone else runs it.

One logistical note, because it decides whether the look stays quiet. Do the recruiter calls and interviews on personal time and personal devices, never the work laptop or the work calendar. Turn off the LinkedIn setting that broadcasts profile edits to your network before refreshing anything, and keep the search off any channel the current employer can see. A quiet-look that stops being quiet, because a recruiter called the wrong person or a profile update lit up the company feed, converts a low-cost information-gathering move into a loyalty conversation nobody wanted. The whole value of the exercise is that the current job situation does not change while the data comes in. Protecting that discretion is what keeps the option open instead of forcing the decision early.

At the end of four weeks, the decision changes shape. It stops being "do I hate this enough to leave" and becomes "given the alternative I now understand, does the current role beat the alternative or not?" That question has an answer. The prior version did not.

What the quiet-look sometimes reveals (and the tripwire that means it stops being quiet)

The quiet-look does not always produce an exit. Roughly a third of the time it produces the opposite. The market read shows the current role is better than the alternatives, and the employee returns to the job with a firmer sense of why they are staying. That is the same process that produces exits, run to a different conclusion.

The other common outcomes: a compelling alternative that justifies leaving (one out of three, roughly), a comp offer that becomes leverage for an internal counter-negotiation (one in five), and a diagnostic that the frustration was about the manager or the trajectory rather than the role itself, unlocking internal moves that were invisible before.

The tripwire that ends the quiet phase is a real offer. Once a real offer at a real number for a real start date is on the table, "quiet" is no longer available. The timeline collapses to the offer's decision window, typically 5-10 business days. At that point the runway math from the I hate my job cornerstone becomes the constraint, and the leaving decision becomes concrete rather than hypothetical.

The job-search-readiness checklist collapses to this: any one green-light signal plus the four-week quiet-look equals enough information to make the leaving decision from data instead of feeling. Waiting for certainty before you look for a new job is the reliable way to leave too late.