One of the most common and widely cited statistics about startups is that 80% of them fail within the first two years of operation.
This stat is often thrown around to show how risky and unpredictable starting a new business can be, and to kind of scare off or caution anyone thinking about becoming an entrepreneur.
But is it even accurate? And where did it come from in the first place? Most important of all, what does it mean for startups and founders?
What defines a failure?
The first challenge in answering this question is to define what constitutes a startup and a failure.
A startup isn’t just any new business—it’s typically something innovative, scalable, and focused on growth. It’s usually built around a fresh idea, product, or service that tries to solve a problem or meet a need in a big, underserved market.
Okay, that might not sound super exciting, but that’s pretty much the standard definition of a startup.
On top of that, startups have the potential to grow really fast—like, exponentially—by using technology, network effects, or other tools like outside funding to scale up quickly.
In more dry terms via HBR: A startup is not a lifestyle business, a hobby, or a side project, but a serious and ambitious endeavour that requires significant investment of time, money, and resources.
One thing we know for certain that beginning a startup is hard. Having it be successful? Well, that's even harder. So, how do we define what a failure is?
On the flip side, failure isn’t just about a business shutting down or going bankrupt—it’s more about not hitting the goals or results the startup was aiming for.
A failure can also be influenced by different factors, such as market demand, product quality, competition, regulation, team dynamics, funding, timing, or luck. A failure can be partial or total, temporary or permanent, and voluntary or involuntary.
So, with these definitions in mind, the next hurdle is tracking down solid, consistent data on startups and failures.
The problem? There’s no one go-to source for this kind of info. Different studies and reports all use their own methods, samples, and metrics, which makes it tricky to compare or trust the numbers outright.
Some of the most commonly cited sources of data on startups and failures are:
- The U.S. Bureau of Labour Statistics: The BLS tracks the survival and exit rates of new businesses in the U.S. based on payroll tax records. According to the BLS, about 80% of new businesses survive the first year, 66% survive the second year, 50% survive the fifth year, and 33% survive the tenth year5. However, the BLS does not distinguish between startups and other types of businesses, nor does it measure the performance or profitability of the businesses.
- The Startup Genome Project: The Startup Genome Project is a research initiative that studies the factors that influence the success and failure of startups around the world. According to the Startup Genome Project, about 90% of startups fail within the first three years of operation. However, the Startup Genome Project does not provide a clear definition of failure, nor does it disclose the sample size or methodology of its research.
- The Harvard Business School: The Harvard Business School is a leading academic institution that conducts research and education on entrepreneurship and innovation. According to a study by Professor Shikhar Ghosh, about 75% of venture-backed startups fail to return their investors’ capital, and about 30% to 40% of them liquidate all their assets. However, the study by Ghosh is based on a small and selective sample of startups that received venture capital funding, which represents only a fraction of all startups.
As we can see, the sources of data on startups and failures vary widely in their definitions, methods, and results. And while the above sources of data are very US centric, we can extrapolate meaning from these regardless, as most startups will face the same difficulties regardless of geography.
Bottom line: it is hard to generalise or compare, and to draw definitive conclusions.
However, based on the available data, we can say that the statistic that 80% of startups fail within the first two years is not entirely accurate or representative of the reality of startups.
A more nuanced and realistic way to approach this question is to acknowledge that the failure rate of startups depends on many factors, such as the type, stage, industry, location, and funding of the startups, as well as the criteria, sources, and methods of measuring failure.
If you're a founder, does this news mean you can relax? Not really. Starting and keeping a business going is hard work. There's still a good chance your venture could fail, but is it really 80%?
What's the Bottom Line?
The question of whether 80% of startups fail is not a simple or straightforward one to answer.
It requires a clear and consistent definition of what constitutes a startup and a failure, as well as a reliable and comprehensive source of data on startups and failures.
Based on the existing data, we can say that the failure rate of startups is high, but not necessarily 80%, and that it varies depending on various factors.
Instead of relying on a single or simplistic statistic, we should look at the broader and deeper context and causes of startup failure, and learn from the successes and failures of other startups.
