What percentage of businesses fail?

By Tom Hill, Editor. Updated November 12, 2026.

Why small businesses fail

No official US source we found records why a business closes today. The BLS survival figures are built from employment records³, so they show that a location stopped employing anyone but not why. The nearest evidence comes from surveys of firms that are still open, which say what they struggle with, and from bank data on how much cash they keep.

What small employer firms say they struggle with

Federal Reserve Small Business Credit Survey, firms with 1 to 499 employees, about 6,500 responses from a weighted convenience sample. Firms that had already closed couldn’t answer, so these are challenges, not reasons for closing. Multiple answers allowed.¹¹

Reaching customers is the top operational challenge, ahead of hiring and keeping staff, and rising costs top the money worries, named by nearly three in four firms¹¹. “Uneven cash flow” includes waiting to be paid by customers¹¹. Getting enough customers and covering costs come up again in the rest of this page.

How many days of cash small businesses hold

In a JPMorgan Chase Institute study of 597,000 small businesses with Chase business accounts, using 2015 data, the median business held enough cash to cover 27 days of its usual outgoings. A quarter held less than 13 days¹⁴. That’s one bank’s customers in 2015, not a national sample, but it shows how short the runway was for most of them.

8 of the 12 industries in the study. JPMorgan Chase Institute, Sep 2016, 597,000 Chase Business Banking customers, Feb to Oct 2015.¹⁴

Restaurants hold the thinnest cushion, about half of what health care and professional services firms hold¹⁴.

The owner’s own money is the backstop

When small employer firms in the Fed survey had financial trouble, the most common response was the owner’s own money: 54% used personal funds, 48% raised prices and 47% drew on cash reserves¹¹. Among firms with no employees, 64% used personal funds¹². For someone starting out, the cushion is the business’s cash plus whatever you can afford to put in yourself.

How many are making money

The Fed survey found that 47% of small employer firms were operating at a profit at the end of 2024¹¹.

The Census Bureau’s Annual Business Survey found that 65.6% of employer firms reported a profit in 2022, a preliminary figure¹³.

The two differ in year, sample and question. The Census survey is a probability sample; the Fed’s is a weighted convenience sample, so the two figures aren’t directly comparable. Either way, a large share of employer firms weren’t making a profit: about half in the Fed survey and about a third in the Census one.

What percentage of businesses fail each year

Each row is a different cohort, all measured in March 2026, which is why years 5 and 6 are almost level. BLS Business Employment Dynamics, Table 7, private sector. “Closed” is 100 minus “still open”.¹

The first-year figure barely moves. Across all 32 cohorts, opened in the years to March 1994 through March 2025, first-year survival has stayed between 75.2% (opened in the year to March 2008) and 80.9% (opened in the year to March 2020)¹.

Following a single cohort gives a smoother line. Of the establishments opened in the year to March 2016, 79.6% were still open a year later, 50.7% after five years and 34.2% after ten¹.

Still open isn’t the same as succeeding, and closed isn’t the same as failed. The section on what these numbers leave out explains why.

Which businesses close most often

BLS Business Employment Dynamics, Table 7 by sector. 1 yr = opened in the year to March 2025; 5 yrs = 2021; 10 yrs = 2016; all measured in March 2026. All private first as the baseline, then sectors sorted by 10-year survival. Seven other sectors, including manufacturing and finance, aren’t shown.²

Of the 12 sectors shown, mining (27.4%) and information (30.5%) have the lowest ten-year survival. Agriculture (51.2%), real estate (44.5%) and retail (43.5%) have the highest².

Accommodation and food services, the sector that holds restaurants, has one of the highest first-year rates at 85.3%, and its ten-year rate of 42.9% is above the 34.2% for all private employers². That goes against the common belief that restaurants close fastest, with two caveats. The sector also includes hotels, and a new branch of a chain counts as an opening⁵. These sector tables don’t show restaurants on their own.

For one business type in depth, see are vending machines profitable?, and every rated type is on the ratings page.

What these numbers leave out

• One-person businesses. The BLS figures come from unemployment insurance records, which exclude proprietors, the unincorporated self-employed and unpaid family members⁴. In 2023 the US had 30.4 million businesses with no employees and 8.4 million employer establishments, so about four in five (78.4%) had no employees^6,7. In 2024 the median one took in between $10,000 and $25,000 a year⁶. The two counts aren’t a clean split: some no-employee businesses are likely side work for people who also have jobs.
• How long those last. The only survival figure we found for them is old: an estimated one third of non-employer firms lasted four years, against about half of new employer firms, in 1990s Census data⁸.
• Locations, not companies. BLS counts establishments, which are based on location⁵. A chain’s new branch counts as an opening, and a closed branch as a closure.
• Closure isn’t failure. About a third of owners whose businesses closed said the business was successful when it closed, in Census data from the 1990s⁹. Owners who started with no capital, or who were young, were more likely to close, and also more likely to call the closure successful⁸.
• The web overstates it. A review of new-venture “mortality” figures found the most-quoted one on the web was “50% after one year”, against official survival in advanced economies of around 80% after one year¹⁰.

Startups are different: the “90%” and “42% no market need” figures

Neither studies small businesses. CB Insights says running out of capital is usually the final cause, not the root one.^15,16

The “42% no market need” figure most pages quote is from 2014, based on 101 post-mortem essays written by startup founders¹⁵. CB Insights’ current study looks at 431 venture-backed startups that shut down since 2023, and there running out of capital comes first, at 70%¹⁶.

A plumber or a café isn’t a venture-backed startup, so neither study says much about one. The “90%” figure is usually said about startups. For new US businesses with employees, BLS shows that 65.8% of those opened in the year to March 2016 had closed by March 2026¹.

How to check yours won’t

Nobody can promise a business lasts, but you can test the two pressure points above, customers and cash, before you spend much.

1. Test demand before you spend. Work through the ten questions on our home page, then do some cheap market research in your own area.
2. Count your cash in weeks. In the Chase study (2015 data), the median small business held 27 days of outgoings¹⁴. Work out how many weeks the business’s cash would cover its costs, and how many months your savings would cover your own bills if it paid you nothing.
3. Imagine it has already failed, and list why. Draper’s guide to running a pre-mortem (made by us) walks through it.

For free help, Small Business Development Centers offer “no-cost business consulting” at nearly 1,000 centers¹⁷. They’re the better place for cash-flow, debt and legal questions.

Draper’s Validate Your Idea workshop (made by us) is free with no card¹⁸ and sits after these, not ahead of them. It helps you list what has to be true for your idea to work, set a test to run before you spend, and decide the point at which you stop. Draper as a whole is about 50 workshops that take you through each step of starting a business, on Draper Start at US$25 a month^19,18.