Don't quit your job on a Sunday night

Quick answer: Stay if you're still learning and your boss would fight for you. Leave if you're asked to compromise your integrity, if the job keeps costing your family more than it pays, or if you asked plainly for change and nothing moved.

It's 9:40 on a Sunday night. The kids are down, the dishwasher's running, and you've got a document open called "resignation_FINAL." You've rewritten the second paragraph four times. Monday opens with a 7:30 standup led by the VP who presented your deck as his own on Thursday, and you're done.

Close the laptop.

You might be completely right. But Sunday night is the worst possible time to make a decision you'll live with for five years. Everything you feel at 9:40 p.m. is true, and almost none of it is the whole picture.

A bad week isn't a bad job

Every job has weeks that make you want to walk out. What you need to know is whether this is a spike or a pattern, and you can't tell from inside the spike.

So keep score. For the next 90 days, take two minutes every Friday afternoon and write a number from one to ten for the week, plus one sentence on what drove it. A note in your phone works fine.

At the end of the quarter, read it back. You'll see one of two things. Maybe the bad weeks cluster around something specific that has an end date, like quarterly close or a project that wraps in June. Or maybe the low numbers show up everywhere, for the same reasons, week after week. The first is a job with bad weeks. The second is a bad job.

The log does something else, too. It turns a feeling into evidence. You'll need that for the next step.

Have the conversation you've been avoiding

Most men who quit never asked for the thing that would've made them stay.

They wanted more money and didn't ask. They wanted off the account from hell and assumed the answer was no. They wanted to stop traveling three weeks a month and figured that was just the job. Then they left, and their manager found out in the exit interview that a thirty-minute meeting could have solved the whole thing.

Before you leave, ask. Be specific. "I'm not happy" gives your boss nothing to work with. "I want to move onto the pricing team by Q3" or "I need travel capped at one week a month because my son starts high school this fall" gives them something they can actually say yes to. If money's the issue, go in with a number, not a mood. Walk into the raise conversation with a number, not a feeling covers how.

Then give it a real window. Sixty to ninety days is fair. Budgets run on cycles, and a boss who says "let me see what I can do" might genuinely mean it.

If the window closes and nothing moved, you have your answer. You also get to leave knowing you gave them a fair shot, which matters more than it sounds like it should.

One warning. Don't make the ask a bluff. If you say "or I'll leave," be ready to leave. The man who threatens and then stays loses leverage he never gets back.

Reasons to stay that don't feel like reasons

Some jobs look dull on paper and are quietly the best thing in your life.

If you're paid fairly and the job lets you leave at 5:15 to make practice, think very hard before you trade that for a 20 percent bump at a startup whose founder texts at midnight. A steady job while your kids are young isn't settling. It's a strategy. You're buying years of dinners at your own table, and you can't buy those back later at any price.

Stay if you're still learning, too. The first couple of years in a new role are usually the steepest part of the climb, and plenty of people bail right before it pays off. If your boss is likely moving up in a year and you're the obvious successor, patience might be the smartest move of your career. The men who get that seat tend to be the ones already acting like it. More on that in The promotion goes to the guy people already follow.

The honest concession: staying too long carries its own risk. Comfort turns into stagnation without announcing itself. If you've done the same job for four years and could do it asleep, "stable" may have quietly become "stuck."

Reasons to leave, no debate

Some things don't need a 90-day log.

You're asked to do something dishonest. Fudging the numbers. Misleading a customer. You don't trade your integrity for a salary, and no bonus makes it smaller. Say no, write down what happened and when, and start looking. If what you're being asked to do might be illegal, talk to an employment attorney before you do anything else.

The job treats your family as an obstacle. Every job has crunch seasons. But if missed birthdays are routine, if taking your daughter to the doctor earns a comment in your review, if your wife now plans around your absence by default, the job is costing more than it pays. No title is worth coming home to people who've learned to do without you. If travel is the heart of it, read The business trip isn't the problem. Coming home is. first, because sometimes that part is fixable.

You've been passed over twice with no straight answer. Once can be timing. Twice, after you've asked what it would take, means they've decided who you're. Go somewhere that hasn't.

You stopped growing, asked to fix it, and got nowhere. That's the log and the conversation doing their jobs. They just returned a no.

Do the money math before the dramatic exit

Unless integrity forces your hand, don't quit without an offer in hand. Searching from a job is easier, and you negotiate from strength instead of need.

Before you sign anything, check what doesn't show up in the offer letter:

• Vesting schedules. Look at your 401(k) match and any stock grants. Walking out a month before a vesting date can cost you real money.
• Bonus timing. Some annual bonuses require you to be on payroll the day they pay out. Read the plan document, not the rumor.
• Health insurance gaps. Find out the exact day your coverage ends and the day the new plan starts. COBRA can bridge it, and it isn't cheap.
• Your runway. Know how many months your household could run on savings if the new thing falls apart.

That's general education, not a plan. For your actual numbers, sit down with a fee-only financial planner and bring your wife to the meeting. It's her risk too.

Leave it better than you found it

How you leave follows you further than the job does. Industries are smaller than they look, and your old boss's opinion of you will come up in a reference call you'll never know about.

Give at least two weeks. Write down everything that lives in your head: where the files are, which client needs a phone call before they'll read an email, why that one spreadsheet has a hidden tab. Train whoever inherits your work, even if nobody asks. Hand your team off in better shape than the day you got them.

Don't trash the company on the way out, exit interview included. Say what's useful, say it calmly, and let the rest go.

And if they come back with a counteroffer, remember why you were leaving. More money doesn't fix a boss who takes credit for your work. If the counteroffer actually solves the problem, great. If it just makes the problem better paid, keep walking.

Decide on a Tuesday

Make the call on an ordinary weekday morning, after coffee, with the log in front of you and the conversation already behind you. Talk it through with your wife before anyone at work hears a word. She's lived this job as long as you have, just from the other side of the front door.

If it still looks like leaving on a Tuesday, go. You'll be walking toward something instead of just away from a VP.

And that file on your laptop? Take "FINAL" out of the name. It isn't final until Tuesday says so.