This site exists for people who choose the grind, so let's be honest about the physics of it. A 72-hour week is not something humans do indefinitely; it is something we can do for a while, on purpose, for a payoff. The people who come out of a 996 stretch ahead treat it like a season with an exit condition. The people it wrecks treated it like an identity. Here is the difference, mechanically.
Define the end before you start
Before day one, write down the condition that ends the stretch: a number saved, a launch shipped, a vest date, a title, a date on the calendar. Two years is a stretch; "until things calm down" is a trap, because things never calm down on their own. Put a checkpoint every quarter — an hour with the written goal, asking one question: is this still buying what I signed up to buy? The goal can change; having no goal cannot.
The floor you cannot trade away
At 72 hours a week you will trade away almost everything discretionary. Three things are not discretionary, because losing them destroys the very output you are grinding for:
- A sleep floor. Seven hours, non-negotiable, most nights. Chronic short sleep quietly cuts the cognitive output the whole schedule exists to maximize — running 996 on six hours is paying for 72 hours and receiving 50.
- The rest day, fully off. Six days is the deal. The seventh day spent "just catching up" turns 996 into 997, and 997 is not a schedule, it is a countdown.
- Twenty minutes of movement a day. Not fitness — maintenance. A walk counts. Twelve sedentary hours a day compound into back, heart, and mood problems on exactly the timeline of a two-year stretch.
Food, social life, hobbies — those you can compress for a season and rebuild. Sleep debt and a missing rest day compound.
Tired versus burning out
Tired is normal here; burnout is different, and catching the difference early is the skill. Tiredness responds to rest — a real weekend off and you want back in. Burnout does not respond to rest, and it announces itself in specific ways: the work you used to care about starts feeling pointless (cynicism is the classic first sign), your output drops while your hours don't, small tasks trigger outsized dread, and you stop being able to remember why the goal mattered. Two honest checks: would you take a promotion at this company right now — and does the written goal still feel like yours? If both answers have gone flat and a full week off doesn't move them, the stretch is over whether you admit it or not. Staying past that point does not earn more; it converts the premium you earned into recovery time you'll spend later.
Ending it well
How you leave determines what the stretch was worth on your resume. A few rules make the exit clean. Decide with your checkpoint, not in the 2am trough — never quit at 2am, and never stay at 2am either; both decisions belong to the quarterly review. Give real notice and ship your handover like it is the last thing they will remember about you, because it is. Check the calendar mechanics before you announce: vest dates, bonus payout conditions, notice periods — leaving three weeks before a cliff is a very expensive way to feel free. And say the true thing on the way out: "this was the deal, the deal is complete" burns no bridges — an employer who posted an honest 996 role understands an honest exit.
Choose it, price it, end it
The whole premise of this board is that intensity should be a transparent trade, not a surprise. That cuts both ways: employers state the schedule in the ad, and you hold up your side by going in with a price (do the math), a floor, and an exit condition. Grind as a chapter, with the ending written first — that is the version that compounds into something.