About Half of New U.S. Business Establishments Survive Five Years

Only about half of new U.S. business establishments are still operating five years later, according to the U.S. Bureau of Labor Statistics. Hong Kong registered a record number of companies in 2025, but the South China Morning Post reported that more than 99,000 were dissolved that year.

Those figures are not directly comparable: the U.S. statistic tracks five-year survival, while the Hong Kong data show annual registrations and dissolutions. The data make failure a measurable operating risk rather than a rare story. Whether it becomes a setback or a turning point depends on the response.

In 1989, I opened an Italian restaurant in Tsim Sha Tsui, Hong Kong, where waiters sang opera arias between courses. The restaurant was fun, but not a good business; it closed in 1993.

I had assumed restaurant operation would be easy. The location was wrong, costs were too high, and my lack of trade knowledge delayed recognition of both problems. Around the same time, I invested in film production and property; both struggled and were sold in the mid-1990s.

Founders often discuss successes and quietly omit setbacks. The cited evidence supports treating failure as a normal part of business, but it does not make recovery automatic.

What should founders know before entering an unfamiliar market?

A 2026 CB Insights analysis of startup shutdowns found that 43% cited poor product-market fit. I suspect many capable founders had misread a market they did not know well enough. No sample size or geography is provided for that figure, so 43% should not be treated as a universal rate.

My rule after those losses is simple: “Never set up a business you don’t know.” In an unfamiliar field, I advise partnering with someone who has spent years in it and listening to that person. The rule comes from experience, not a controlled test, but it addresses knowledge gaps before capital is at risk.

Takeaway: Test the market before committing capital, and seek operating knowledge you do not have.

When should a struggling business close?\n

CB Insights found in the same analysis that 70% of failed startups ultimately ran out of capital. I treat that as a last symptom, not the cause; when cash runs out, months or even years may have been lost. The statistic does not show how many of those failures were preventable.

Crowdfund Insider, reporting on the same analysis, said nearly a quarter had been “walking dead” for three years or more after their last funding round. They remained in operation but no longer grew. Hanging on rarely saves such a business and more often makes the eventual decision more expensive.

Selling the film and property businesses was painful, but it freed time, money and attention for something better. Closing a venture is not the same as failing as a person.

Takeaway: Do not wait for an empty bank account to recognize that a venture has stopped growing.

How should a founder recover without losing judgment?\n

Professor Dean Shepherd’s work on grief and business failure, published in the Academy of Management Review, argues that founders recover and learn faster when they move between confronting the loss and stepping away from it. The work rejects both dwelling indefinitely and pretending the loss never happened, but it offers no fixed recovery timetable.

A study of 242 entrepreneurs led by Dr. Michael Freeman of the University of California, San Francisco, found 72% reported being affected by mental health conditions, either directly or through their families. The 72% figure is broader than a clinical diagnosis and does not establish that business failure caused those conditions.

The practical advice is to talk to someone trusted, sleep properly and keep exercising so the next decision is made with a clear head.

My father died when I was 18, aged just 49, and that loss made me take my health seriously long before it was fashionable. In difficult business periods, that habit has helped keep me steady.

Takeaway: Protect your health and decision-making before deciding what to do next.

What should a failed venture teach the next one?\n

Harvard research by Paul Gompers and colleagues found a 20% next-time success chance for venture-backed entrepreneurs whose previous company had failed, versus 18% for first-time founders and 30% for those with a prior success. The failure group therefore sat only two percentage points above first-time founders. No sample size or significance test is provided for those figures, so the gap is directional, not a guarantee.

Harvard Business School professor Amy Edmondson’s Right Kind of Wrong separates “basic failures,” caused by avoidable mistakes, from “intelligent failures,” which test a sensible idea in new territory while keeping the downside small. The first kind should be prevented; the second is the price of progress.

After a setback, I write down what I assumed, what actually happened and what I would do differently, then ask someone outside the business to challenge it. Experience alone does not make a better founder; practice turns an early judgment on ability into usable information.

Takeaway: Turn a failure into a documented test of what to change, not a verdict on your ability.

Does trying again quickly improve the odds?

A 2019 Nature study by Yian Yin, Dashun Wang and colleagues analyzed more than 58,000 venture-backed startups alongside other fields. People who eventually succeeded did not fail less often than those who did not. Their new attempts came faster because they built on what they had learned instead of starting from scratch.

That finding does not show that every rapid restart works or quantify a success gain from speed. Speed here does not mean recklessness; it means shortening the gap between learning a lesson and acting on it.

The Global Entrepreneurship Monitor’s 2025/2026 report found that fear of failure deters two in five adults from starting a business. It also found that people who had previously exited a business were more likely to start again, though the report gives no size for that advantage. The pattern supports another attempt more than it guarantees one.

As I once told an interviewer, “all these failures are what have shaped me.” My advice to a founder recovering now is to be patient, be honest about what went wrong, and go again when ready.

Takeaway: For a founder deciding whether to restart, the next move is a measured launch built on documented lessons, not speed for its own sake.