After a career break: should you get a job or start a business?

If you have been out of the workforce for a few years, for children or anything else, the decision comes down to three things: how much money you need and by when, how many uninterrupted hours you actually have in a week, and what happens if nothing works after six months. If you need a specific amount by a specific date, a job is the safer answer. If you have some runway and what you want most is control over your time, starting something is a reasonable bet. Most people cannot answer those three questions yet, which is why the decision feels stuck. The rest of this page is how to answer them.

One thing to know before you read on. Almost every guide you will find on returning after a break is published by a company that sells one of the two answers: a job board, a returnship program, a coaching package. We build a tool that helps people start something, so we have our own bias. We have tried to write the version we would want to read, including the parts that argue against us.

How much money do you need, and by when?

Not "how much would be nice". The number you must cover, and the month you must cover it by.

If that number is due in under three months, take the job. Starting something almost never produces reliable income in that window, and a business that has to pay a bill next month makes bad decisions: wrong customers, prices too low, work you cannot sustain.

If your number is smaller than you think, or the date is further out because a partner's income or savings covers the floor, then you have runway, and runway is the thing that makes starting possible. Be exact about how many months it is. "Some savings" is not a plan.

Watch out for the trap in the other direction too. A job has costs that eat into the number: childcare during fixed hours, commuting, the days you cannot take off. Compare what you would keep, not what you would earn. For plenty of parents that math has already been run, and it ended the discussion:

Plus, my entire paycheck would go to daycare for two kids so it doesn't even make sense for me to go back.

From a thread on r/SAHP.

How many uninterrupted hours do you actually have?

Count them for one real week before you decide anything. Not the hours you hope to protect, the ones you actually had.

This matters more than motivation. Most of the work of starting something can be done in short sessions, but some of it cannot, and if your week is made of fifteen-minute fragments you need to choose something built for fragments: a service with a fixed scope, a product you sell the same way every time, a business that does not need you online at a specific hour.

Because the deeper problem is not your week. It is that jobs mostly come in one shape, and your week no longer fits it. Here is what that costs, in the words of a parent who decided to retrain rather than return:

The type of jobs I would want don't have part-time positions, so when I was considering a career change, the flexibility of being part time was the most important thing to me. So I'm back in school in the healthcare field learning phlebotomy. I think it is beneath my current work experience, but I think I'll be more likely to find a part-time job.

From a thread on r/SAHP.

Read that again: she is accepting work beneath her experience to buy flexibility. That trade is worth knowing about before you make it, because it is the strongest argument in this guide for at least pricing the other path, where flexibility is the default rather than the thing you pay for with your CV.

If your honest answer is under five hours a week, that is not a reason to give up. It is a reason to pick something smaller than the thing you are imagining.

What happens if it does not work in six months?

Write down the answer. Then check whether you can live with it.

For most people the honest answer is: I would look for a job, six months later than I would have otherwise. That is a real cost and it is worth naming. Research on hiring suggests it is smaller than people fear, though. In a field experiment sending roughly 12,000 applications to real openings, callback rates dropped as a jobless spell got longer, but the drop flattened out after about eight months, rather than continuing to fall (Kroft, Lange and Notowidigdo, Quarterly Journal of Economics, 2013). A longer gap does not appear to be proportionally worse than a shorter one.

What matters more is what you did with the time, and how you explain it. In a study of 163 recruiters assessing 815 candidate profiles, a break taken to care for family scored better than a break spent unemployed, and a break spent studying or training scored substantially better than both (D'hert, Lippens and Baert, Ghent University working paper, 2024; the study was run in Belgium with recruiters assessing profiles, not real applications, so treat the sizes as indicative).

There is a practical version of this. In a preregistered experiment with 9,022 real job applications in the UK, listing total years worked instead of employment dates increased callbacks, and the effect was largest for CVs that had a gap (Kristal, Nicks, Gloor and Hauser, Nature Human Behaviour, 2023). Six months building something of your own is years worked. It is also a specific thing to talk about in an interview, which is more than most people come back with.

What does the job path actually look like right now?

Better than the internet suggests, and harder than it was.

As of July 2026 the US unemployment rate was 4.1%, with 6.9 million people unemployed. Of those, 1.8 million had been out of work six months or longer, a quarter of the total (BLS, Employment Situation, August 2026). So people are being hired, and long searches are common enough that yours would not be unusual.

Two honest points in favour of the job.

Returning sooner is worth real money over a career. Work looking at mothers' earnings found that shortening a non-employment spell by one year was associated with roughly 4.6 to 6.8% higher wages ten to twenty years later (Kuka and Shenhav, Federal Reserve Bank of San Francisco Economic Letter, October 2023). Delaying a return is not free.

Returnship programs work for the people who get into them. Path Forward states that more than 80% of returners who complete a returnship convert to a full-time position. Read that number carefully: it describes people who were accepted into a program and then finished it, so it is not the odds you face from the outside, and it is self-reported by the organisation that runs the programs. Still, if you get a place, it is a strong path and we would tell you to take it.

What does starting actually cost you?

Four things, in roughly this order, and none of them are the fun part.

Setting up: registering something, a bank account, invoices, working out what you owe and to whom. Unglamorous, finite, and the thing that stalls most people for months.

Making the thing: the product or the service, in a form someone can actually buy.

Finding out whether anyone wants it. This is where most projects quietly end, and it is not a step you can skip by making the thing better.

Selling it, then doing it again next month. The first customer is a good day. The tenth is a business.

Two numbers worth holding in view. About 22% of new US business establishments close within their first year, and roughly half are gone within five (BLS, Business Employment Dynamics, survival tables, data through March 2025). Those figures count closures, which includes planned wind-downs and sales, not only failures. They have also been remarkably stable across cohorts. Read them as: this is normal and survivable, not as: four out of five make it.

And you are not unusual for wanting this. 62% of US adults say they would prefer to be their own boss, against 35% who would rather be an employee, in a Gallup survey of 46,993 adults (Gallup, June 2024). Wanting it is common. Doing it is the part that is rare, and the reason is usually the four steps above rather than a lack of ideas.

Can you do both at once?

Usually, and often you should.

Look for work while you build, and let each one inform the other. A job you take now does not close the door: plenty of businesses were started by someone with a salary, and a salary is the cheapest runway there is. Equally, a small thing you sell to a real customer while job hunting is the most concrete item you will have to talk about in an interview.

The version that does not work is doing both at 30%, on the theory that you will decide later. Pick which one gets your best hours this month, and give the other the leftovers on purpose.

So which one?

A rough rule, from the three questions.

Take the job if your number is due within three months, or if the honest answer to "what if it does not work" is something you cannot live with.

Start something if you have at least six months of floor covered, you can name five uninterrupted hours a week, and what you want most is control over when you work rather than a specific salary.

And if you are in between, you are in the biggest group of all. It sounds like this:

If your salary is more than the cost of daycare, then technically you wouldn't be "saving", but that doesn't mean it's not worth it. [...] So I need to go back part-time when he starts school but I also want to have another baby in 2-3 years from now. Trying to figure out how to make it all work lol.

From a thread on r/stayathomemoms.

If that is you, do the smallest version of starting that fits your real week, while you look for work. You will learn more in a month of trying to sell something than in another month of deciding.

One more option, briefly

If the four steps above are what stops you, and specifically the first one, that is the gap we built baget for. You describe the idea in a sentence and an AI team sets up the site, the copy and the way customers pay, in sessions short enough to fit a real week. Finding the first customer and doing the work well is still yours, and that was always the hard part.

Seeing what your version would look like costs nothing.

Sources

• Kroft, Lange and Notowidigdo, Duration Dependence and Labor Market Conditions, Quarterly Journal of Economics, 2013: field experiment, roughly 12,000 applications. academic.oup.com
• D'hert, Lippens and Baert, Ghent University working paper 2024/1100: vignette study, 163 recruiters assessing 815 profiles, run in Belgium. wps-feb.ugent.be
• Kristal, Nicks, Gloor and Hauser, Nature Human Behaviour, 2023: preregistered experiment, 9,022 real job applications in the UK. nature.com
• BLS, Employment Situation, July 2026 data released August 2026: unemployment rate, long-term unemployed share. bls.gov
• Kuka and Shenhav, How Much Do Work Interruptions Reduce Mothers' Wages?, FRBSF Economic Letter 2023-25, October 2023. frbsf.org
• Path Forward, returnship FAQ: the 80% conversion figure, self-reported. pathforward.org
• BLS, Business Employment Dynamics, survival of private-sector establishments, Table 7, data through March 2025. bls.gov
• Gallup, June 2024, survey of 46,993 US adults: preference for being one's own boss. news.gallup.com