Learn
Should I quit my job to start a startup?
October 8, 2026
A decision framework for leaving a job to start a company: runway, evidence, the cost of staying, and a test you can run before you resign.
"Should I quit my job to start a startup?" is usually two questions taped together. One is about the company. The other is about escape. They deserve different answers, and mixing them is how people resign into a problem they have not defined.
A job can be the wrong container and still be the income that makes a company possible. A company idea can be real and still be too early to carry your rent. The decision is a framework, not a mood you wait to feel resolved.
The question under the question
Before the spreadsheet, say which sentence is truer.
"I want this specific company to exist, and I am the person who has to build the next piece of it."
"I want out of this job, and a startup is the most respectable door."
Both can be present. If the second sentence is doing most of the work, solve the job problem on its own terms. A different employer, a smaller role, a leave, or a career path you have not tested yet may be the actual move. Starting a company to avoid a bad manager is an expensive way to change managers. You become the manager, and you still have the original restlessness at 11 p.m.
If the first sentence is doing the work, keep going. The rest of this page is for that case.
A framework with four checks
Use four checks, in this order. Skip none of them because one feels flattering.
Runway
Runway is the number of months you can pay your real life if revenue stays at zero. Real life means the rent or mortgage you have, the health insurance you need, the debt you service, the people who eat because of this salary, and a plain buffer for the car, the tooth, the flight home. It does not mean the budget you invent the weekend you are angry at your boss.
Write the number down. Then cut it by the months you are not willing to live at that level. A runway you resent by month three is not a runway. It is a countdown you pretend is a plan.
There is no magic threshold that makes the choice safe. A longer runway buys a calmer test. It does not buy a good company. A short runway can still be rational when revenue already exists and the job is the thing blocking delivery. It is rarely rational when revenue is a hypothesis.
Evidence
Evidence is contact with the work of the company, not contact with the idea of it. Talking to friends does not count. A deck does not count. A domain name does not count.
Evidence looks like this: a stranger, not a friend doing you a favor, has paid, signed a pilot, or pulled your unfinished thing into their actual week more than once. Or you have personally done the core operation, the unscalable version, enough times to know the Tuesday shape of it.
If you have opinions and no evidence, you are early. Early is a legitimate place to stand. It is a bad place from which to burn the salary.
Cost of staying
Staying has a cost, and people who love their job title tend to ignore it. So do people who are afraid.
The cost is concrete. Daytime hours the company needs and the employer owns. A skill that dulls because the job no longer uses it. A window with a buyer that closes if you answer in three weeks instead of this week. An integrity problem, if the employer claims work you are building on the side, or if the contract forbids it.
"I am bored" is a weak cost. Boredom is solved by harder work in several directions, including inside the job. "The pilot customer needs me on Tuesday at 10, and I am in a planning meeting" is a real cost. Name the cost in a sentence a skeptical friend can check.
Reversibility
Ask how hard it is to come back to work like this if the company fails or you shut it down on purpose. Some careers forgive a year away. Some treat a gap as a demotion you have to argue out of. Some licenses and client relationships decay on a clock.
Reversibility is not a reason to stay forever. It is a reason to leave with a written return path: who you can call, and what proof of work you keep fresh. People who leap and also burn the bridge are often protecting an identity, not making a business decision.
Two pictures of the same question
Priya leads operations at a mid-size company. She wants to start a meal-planning service for new parents. She has about four months of savings, a partner with a stable job, and no customers. Her notes app is full of brand ideas. She has cooked for friends and collected compliments.
The framework says the resignation is early. The partner's job changes the household risk. It does not create evidence. Priya can sell ten paid plans on nights and weekends, with a plain offer and a delivery date. If ten strangers will not buy, quitting does not produce the tenth customer. It only removes the salary that makes a slow tenth affordable. Her next move is the offer, not the resignation letter.
Andre has eighteen months of expenses, a co-founder who already left their job, and three companies paying for a narrow pilot. The pilots are stalling because Andre can only work them at night, and the buyers work in the daytime. The product is clumsy and the buyers are still renewing the pilot. Staying is now the expensive choice. The cost is not boredom. The cost is a living deal going quiet.
Andre's framework points toward leaving, with a kill date written before the last day of work: a date by which the pilots convert to a defined amount of revenue, or the company stops and the job search starts while the runway is still dull rather than desperate. A kill date is a decision you make calm, so you do not have to invent one while frightened.
A third structure between stay and leap
Build while employed, on a clock. Give the side test a fixed number of weeks. Protect the hours. Stop if the contract forbids the work, and do not steal the employer's time to prove a point. At the end of the clock, you have evidence or you have a no. Both are useful. An open-ended side project is a hobby that taxes your sleep.
Join instead of found. If what you want is the early-company problem and not the founder problem, those are different jobs. Joining a small company as the first operator, designer, or engineer can scratch the same itch with someone else carrying the payroll risk. Be honest if status is the reason you insist on being the founder. Status is a costly reason.
Change the job and keep the idea in a drawer. Sometimes the idea is good and this year is the wrong year. A less consuming job can fund a later start better than a heroic exit followed by a crawl back. Parking an idea is allowed. For how to tell a parked idea from a dead one, see which startup idea is right for you.
What to have on paper before you resign
If the four checks point toward leaving, write this down before you tell anyone at work.
The offer, in one paragraph a buyer understands. Who it is for, what they get, what it costs, and what you do every week to deliver it.
The runway, in months, using last month's spending, not a cleaned-up version.
The kill date, and the number that counts as "this is working" by that date. Pick a number you cannot wiggle: paid revenue, signed pilots, or a waitlist of people who have already given you something costly, like time or money. Vanity metrics are how kill dates dissolve.
The return path. Three people you can call, and the skill you keep fresh so it stays hireable.
Who is affected at home, and what they have actually agreed to. A nod during an excited conversation is not agreement to uneven income. Have the plain conversation, and include the kill date.
What this decision cannot do
A company is a set of repeating tasks that other people value enough to pay for. If you want that, the resignation is a scheduling decision: when your hours move from the employer to the buyer. If you want a different life and the company is the costume for it, settle the life question first. Otherwise you recreate the old stuckness with worse insurance.
If you want an outside document on the chapter you are in before you make a move this large, Meridian Reading is a private 22-section life-path reading, reviewed individually and delivered within 48 hours, for $197 once. It does not tell you to quit. It gives you a structured mirror of how you are built and what this season is, which is a different input from a pep talk. The category is what a life path reading is. How it differs from an astrology reading, and what to expect if you order one, are in life path reading vs astrology reading and what to expect from a life path reading.
See it on the page
Explore the full sample reading in your browser, or begin your own Entry reading when you are ready.
FAQ
Common questions
How much money should I have saved before I quit?
Can I start the company while I still have the job?
What if the idea is only half formed?
How do I tell fear from a real constraint?
Related
When your work stops feeling like you
When craft and identity fuse, skill starts to feel like a cage. How to keep the craft and change the name you answer to.
Life path reading vs astrology reading: what's the difference
An astrology reading stays in one tradition. A life path reading is a wider document about your chapter. Here is how to tell them apart before you pay.
What to expect from a life path reading (and who it's for)
What a life path reading includes, who it helps, what it refuses to decide for you, and how to judge the document before you pay.