The Pivot Doesn't Fail in Month One. It Fails in Month Six.

Most career pivots don't die at the start. They die around month four to six, when the excitement has worn off and you've finally learned enough to see how much you don't know. The dip you feel then is not proof the move was wrong. It's the predictable middle of every real transition. People who quit in month five usually quit a week before it gets better.

That's the part nobody warns you about. The advice industry front-loads everything onto the decision and the landing: pick the right path, polish the resume, get the offer. Then it goes quiet exactly when you need it, in the months when the gap between your old competence and your new domain is widest and the old you is yelling the loudest. So you misread normal difficulty as a signal. You treat a phase like a verdict.

Why does the pivot feel worst around month six, not month one?

Because the first stretch runs on novelty, and novelty has a shelf life. Researchers studying voluntary job changes found a "honeymoon-hangover" pattern: job satisfaction tends to peak right after the move, then declines over the following months, measured across the first year in a sample of managers who changed jobs (Boswell, Boudreau & Tichy, Journal of Applied Psychology, 2005). The high at the start isn't a reliable read on the decision. It's chemistry. And it fades on a schedule.

What replaces it is competence math. Michael Watkins, surveying CEOs about new leaders, put the average break-even point, the moment a midlevel leader gives back as much value as they consume, at 6.2 months, based on the estimates of 210 CEOs and presidents (Watkins, The First 90 Days, introduction excerpt). That figure is leadership-specific, so don't stretch it to every career. But the shape generalizes: for the first several months in a new domain, you are net-consuming. Learning. Making mistakes. Needing help. For a career changer, that stretch is longer and you feel it more sharply, because you used to be the person others came to.

What does the data actually say about early exits?

It says two different stories are hiding in one number, and you have to separate them.

The raw attrition is real. More than a third of newly hired employees quit within their first year (Work Institute, 2024). One onboarding study found 31% had quit within six months, with the exit rate holding at roughly 16 to 17% across each of the first three months, about 1 in 6 new hires per month (BambooHR, The Definitive Guide to Onboarding). Jobvite's seeker research put 25% of workers who left a job as having gone within the first 90 days of starting the new one (Job Seeker Nation 2026).

Look closely and the early cluster is a different animal. People leaving in month one to three are mostly bad fits: the role was misrepresented, the manager is a problem, the job was nothing like the pitch. Those are correct exits. Leave fast.

The month four to six exits are the heartbreakers. These are people who made a sound decision and then walked out during the hard middle, reading the difficulty of learning as evidence of a mistake. Same statistic, opposite truth. One group should run. The other is one bad week from the part where it works.

Why does the old job suddenly look so good in month five?

Because you're stuck between two identities, and the old one doesn't go quietly.

Researchers call this the liminal phase: you're "betwixt and between," disengaging from one identity without having fully left it while taking on another without having fully assumed it (Ibarra & Obodaru, 2016). It's uncomfortable by design. The discomfort is the old self dissolving, not the new one failing. And the pull tends to feel strongest for people with sharp, specific professional identities, the ones who were genuinely good at the thing they left. The more your old self was defined by being competent, the more its undoing stings, and the easier it is to let the years you already invested talk you into staying somewhere they're the reason to leave.

This is why the data looking bad and you stalking your old colleagues on LinkedIn happen in the same week. The old identity isn't abstract. It shows up as a former peer's promotion in your feed at 11pm, a competitor's salary thread, the title you used to have sitting next to someone else's name. That's not a sign to go back. That's the mechanism working exactly as described.

Isn't the "valley of despair" just a coaching cliché?

Partly, and let's be honest about that. The "valley of despair" curve you've seen in slide decks is usually drawn from the Dunning-Kruger effect, and that model is contested. Critics argue the famous shape is largely a statistical artifact of regression to the mean rather than a real trajectory of confidence (Fix, The Dunning-Kruger Effect is Autocorrelation, 2022). So don't lean on the cartoon.

You don't need it. The honeymoon-hangover research is peer-reviewed and measures the thing directly: satisfaction high at entry, declining over the following months (Boswell et al., 2005). The phenomenology holds even where the pop-science curve doesn't. People in the middle of learning a hard thing report feeling worse about their ability than they did at the start, and worse than they will at the finish. Keep the experience, drop the bad chart.

How do I tell the valley from a genuinely wrong move?

This is the question that matters, and "just push through" is dangerous advice if the role is actually broken. So separate the two honestly.

Signal	The valley (stay, keep going)	A wrong move (leave clean)
What you can name	You can finally describe what you don't know yet. Three months ago you couldn't	The problem is the people or the lie, not your own gap
Where the pain lives	Domain-specific. The work is hard	Identity-corroding. You're being demeaned, set up to fail, or sold a fiction
Motion	Slow forward progress. Small wins exist if you look	Flat or backward. No path to wins regardless of effort
The manager	Stretching you, occasionally badly	Misrepresented the role, or is genuinely abusive

The clean rule: the valley is about the gap between your old competence and the new floor. A wrong move is about the role itself being broken or dishonest. Knowing the dip is coming should never become a reason to endure a bad situation longer. It's permission to stop reading normal difficulty as a verdict, nothing more.

What does this look like with a real person?

Weak read: "I knew product management wasn't right for me. I was struggling in month five and nothing clicked." That's the gap, misread as a verdict.

Strong read: "In month five I could finally see how much I didn't know. The map of my own incompetence was terrifying, and it meant I was learning. Six months later I led my first launch solo." Same five months. The clarity of the gap was evidence of growth, not failure. This is also why the sense that the work finally fits tends to arrive after the competence does, not before: the conviction is downstream of the grind, so its absence in month five tells you nothing.

Take the engineer who moves into people management. Months one to three: energized, praised for her technical credibility, still solving problems the old way. Months four to six: the team's frustrated, she can't code her way out of a conflict, and "the technical expert" identity she built over a decade feels under attack. She nearly reverts to IC work. What she's in is not failure. It's the distance between her old peak and her new floor.

Or the senior lawyer who joins a startup product team. First 90 days, valued for precision and clear thinking. By month five, she's shipped nothing, paralyzed by ambiguity, and her friends who stayed in law are ahead on title and pay. The old self isn't a feeling here. It's a salary thread at 11pm.

One Praxy user who pivoted from finance to UX described month six as "the month I nearly talked myself out of it three times." Three courses, two portfolio projects, still felt like an imposter. Month nine, first UX role. The valley was real. So was what came after.

What's the trade-off in staying through the dip?

Time and earnings, said plainly. Staying through the valley means months of feeling behind, often at a comp setback, while peers who didn't move compound their old advantage. That's the bill, and every pivot comes with a pay cut whose only real question is how long it lasts. The career-changer's version of Watkins' break-even is steeper: your consumption is higher and your awareness of it is sharper, because you can feel yourself not yet earning your salary in the new domain.

The honest framing isn't "it'll all work out." It's: that setback is the price of the optionality you bought. If the deep want underneath the pivot is real, recognition, autonomy, a kind of work you actually care about, the price is usually worth it. If you can't name the want and you just wanted out, the valley will expose that. Either way, you want to find out on purpose, not by quitting in a panic in month five.

What to do now

• Name the valley before you enter it. Write down, today, that months four to six will feel like a mistake and that the feeling is scheduled, not diagnostic.
• Set a break-even target. Borrow Watkins' idea even if you're not a leader. Define two or three early wins that would prove you're contributing, and a rough date. Track against that, not against your old peak.
• Separate signal from noise weekly. Each week, ask the table's questions. Is this the gap, or is the role broken? Write the answer down so month five can't rewrite month one's decision.
• Don't decide in the dip. Experiment in it. When you want to quit, run a small test instead, a project, a conversation, a week of doing the new work differently, before you touch the escape hatch.
• Protect your inputs at 11pm. The LinkedIn stalking is the old identity talking. Mute the feed for the hard months.

The dip is coming. That's not a warning. It's the one piece of information that turns month five from a verdict into a phase.

Heading into the hard middle of a pivot, or already in it? Message Praxy on WhatsApp. Tell me where you are in the transition, and I'll help you read whether this is the valley or a genuinely wrong move, and what your next early win should be.