Why did 500 startups shut down?

Our Real World Problem

Over 85% of companies worldwide won't survive up to 10 years.

Just in the USA alone, more than 360,000 companies shut down every year (U.S. Small Business Administration, 2019).

Without a doubt, the rate of companies shutting down every year is staggeringly high.

Abstract (TL;DR)

Objectives: Every year around the world, billions of dollars were lost due to business failures. Therefore, the purpose of this analysis was to gain a better understanding on why companies from around the world shut down.

Methods: This was a descriptive analysis conducted with secondary data gathered from the Internet. Due to the current small amount of verified and cleaned data, only 500 inactive companies were observed for this analysis.

Findings: Out of 500 companies, 59 (11.8%) companies had no product-market fit advantage, 75 (15%) companies suffered from poor business models, and 96 (19.2%) companies faced strong competition. Respectively, 73 (14.6%) companies shuttered due to the lack of funds, while 40 (8%) companies from failed fundraising. The shutdown rate was high among 131 (26.2%) startups, 141 (28.2%) small companies, and 82 (16.4%) SMEs after they had secured seed, early stage, and M&A funding. Non-innovative new and small companies generally shut down within 1-3 years (11 companies at 2.2%) and 3-5 years (26 companies at 5.2%) of operation. Overall, the 385 (77.2%) companies that operated in technology related industries were predominantly co-founded by 821 (87.5%) males founders.

Conclusions: A majority of the reasons for shutdown are technical shortcomings which could be avoided through experiential learning, coaching, and mentoring. This report implicates that failure may be part of the process of building a successful company, especially an innovative one.

Introduction

Between 2016 and 2018, our global startup economy created USD $2.8 trillion in value, a 20.6% increase from 2015 to 2017, and it continues to grow every year (Startup Genome, 2019).

Evidently, many research and data have shown that entrepreneurship have positive effects on employment creation, innovation, and economic growth in our global economy (Audretsch & Fritsch, 2002; Baptista, Escária, & Madruga, 2008; Carree & Thurik, 2010).

In achieving the Fourth Industrial Revolution, entrepreneurs are leading the way as agents of change, bringing new ideas to the markets, and driving growth through their competitive advantage (Wong, Ho, & Autio, 2005).

However, most business studies and entrepreneurial research seem to have a natural tendency to focus on success stories (Madsen & Desai, 2010), and less on failure stories which may result in a survivorship bias that can lead to over- or understating the predictability of events (Brown et. al., 1992). Consequently, with the lack of in-depth research on business failures, how can we effectively learn from failures to reduce costly mistakes and avoid poor strategies?

Since our inception in September 2017, Flipidea (pronounced as 'flee-pee-dia') is a Data-as-a-Service (DaaS) platform offering analytical information about business failures. In other words, we analyse business failures.

For this reason, we continuously gather information on companies that ceased operations to examine why they shut down, and draw meaningful insights from our regular analysis.

Our ultimate mission is to help our audience build successful businesses by making data-informed decisions and prudent strategies.

This report is divided into the following sections:

• purpose for this descriptive analysis
• data collection process and its quality
• research questions and methodology
• findings based on the descriptive analysis of 500 inactive companies
• results and discussion
• limitations and future research

Our Purpose

As our data grows, the overarching purpose of our descriptive analysis is to scientifically examine the datasets and gain a better understanding on why companies from around the world shut down. The objectives are:

• to study the reasons for shutdown
• to discover meaningful patterns in the data
• to contribute our findings to existing knowledge and research literature on business failure, entrepreneurship, venture capital and private equity investment, management, and innovation

Our Data

The report is based on secondary data we had gathered from the Internet.

The data are combined datasets which are publicly available on the Internet. Our data retrieval systems identified the failed companies and gathered the data from the companies' websites, blogs, social media, news articles, media interviews, research papers, analytical reports, and so on.

Overall, the datasets comprised of company information, post-mortem reports (an analysis of an event after it is over), investment data, business performance data, founders' profiles, social media data, and so on. However, there are substantial amount of missing data in our data gathering because many of the companies did not publish the information.

Although we take reasonable measures to ensure that our gathered data is accurately reflected in this report, we do not warrant the completeness or accuracy of data provided because our data retrieval systems scan the Internet to identify, monitor, and gather relevant, aggregated, and public information which may be incomplete or inaccurate or not available. Hence, we encourage you to independently verify the accuracy of the information.

In addressing this issue, we leverage on our human-in-the-loop process to verify and manage the integrity of our data while we improve our data retrieval systems, and gradually publish the verified and cleaned data unto our platform. Thus, data in this report is subject to change without notice.

Finally, when interpreting the data, it is important to keep in mind that our datasets include information of companies from around the world, so the data should be interpreted in such context.

Our Methodology

Our descriptive analysis seeks to obtain insights from the 500 inactive companies in our live database, which are also published on www.flipidea.co. Therefore, we only observed these 500 inactive companies for this report.

Firstly, what is descriptive analysis? Descriptive analysis provides information on the basic qualities of data and includes descriptive statistics, such as range, minimum, maximum, and frequency. It also includes measures of central tendency, such as mean, median, mode, and standard deviation. Therefore, it is important to note that descriptive statistics merely describe the observed data.

Since our data are classified (systematic arrangement in groups) and stored in numerous collections, we wrote scripts (aka scripting language, which is a programming language to automate the execution of tasks) to calculate our descriptive statistics.

Due to our current small amount of verified and cleaned data, we merely conducted descriptive analysis to understand the following:

• what are the top reasons for shutdown?
• where did the companies base at?
• which industries did the companies operate in?
• what were the lifespans of the companies?
• what were the last funding statuses of the companies?
• what were the total funding amounts raised by the companies?
• how many employees did the companies hire?
• what is the gender statistics among the companies' founders?

For all of the descriptive statistics, the percentages were calculated based on 500 inactive companies and the frequency (n) of each data features or variables.

percentage % = (n/500)*100

For the gender statistics, the percentages were calculated based on 938 data objects (which comprised of 903 founders, 11 corporations, and 24 no data) and the frequency (n) of each data features or variables.

percentage % = (n/938)*100

In the event you wish to calculate the frequency (n) of each data features or variables, based on the calculated percentages:

• gender statistics: base on 938 data objects
• the other descriptive statistics: base on 500 inactive companies

n = (calculated percentage%/100)*[500, or, 938]

For funding status, we currently classified the different financing stages as:

• seed: angel, pre-seed, seed
• early stage: pre-series A, series A, series B
• late stage: series C and onwards, mezzanine
• IPO (Initial Public Offering): fixed price offering, book building offering

In order to help our audience understand the terminologies and definitions commonly used by entrepreneurs and investors in the tech startup and business scenes, we put together a glossary of business failures.

The glossary is a compilation of terminologies used in business, finance and tech industries that were defined by experts found on the Internet.

Lastly, we classified the companies into different categories by the number of employees in accordance to the criteria of Organisation for Economic Co-operation and Development (OECD, 2020):

• startups or micro-enterprises: lesser than 10
• small enterprises: 10 to 49
• medium enterprises: 50 to 249
• small-and-medium enterprises (SMEs): lesser than 250
• large enterprises: 250 or more

Our Findings

We present the results of our descriptive analysis in this section.

What are the top reasons for shutdown?

From our classification of reasons for shutdown, the post-mortem data revealed that a company may shut down due to multiple reasons with the top 5 common reasons listed below:

1. strong competition experienced by 96 companies at 19.2%
2. poor business model experienced by 75 companies at 15%
3. lack of funds experienced by 73 companies 14.6%
4. no product-market fit experienced by 59 companies at 11.8%
5. failed fundraising experienced by 40 companies at 8%

From the observation of the top 5 reasons for shutdown, the outcome of lack of funds and failed fundraising could seemingly be related to poor business model design, no product-market fit advantage, and strong competition.

Where did the companies base at?

From our classification of data, the top 5 locations are:

1. 236 companies at 47.2% were based in United States
2. 115 companies at 23% were based in India
3. 28 companies at 5.6% were based in United Kingdom
4. 12 companies at 2.4% were based in Canada
5. 10 companies at 2% were based in Indonesia

The data showed that the companies were mostly from North America, India, Europe, and Southeast Asia. There are two main reasons:

1. the Total early-stage Entrepreneurial Activity (TEA) rates in North America (United States 17.4%, Canada 18.2%), Europe (United Kingdom 9.3%, Germany 7.6%), and India (15%) are substantially high (Global Entrepreneurship Monitor, 2020)
2. currently, our data retrieval systems only monitor companies from all English-speaking countries, but we are gradually populating our database with data from non English-speaking countries as well

Which industries did the companies operate in?

From our classification of data, the top 5 industries are:

1. 289 companies at 57.8% were classified as internet industry
2. 25 companies at 5% were classified as blockchain industry
3. 18 companies at 3.6% were classified as retail industry
4. 17 companies at 3.4% were classified as food & beverages industry
5. 16 companies at 3.2% were classified as consumer electronics industry

The data showed that 385 (77.2%) companies operated in the technology related industries, such as internet, blockchain, consumer electronics, transportation, information technology & services, interactive media & services, media, renewables & environment, machinery, and application software.

What were the lifespans of the companies?

From the timeline of companies' earliest formation date to the latest shutdown date, we recorded July 1841 to 31 May 2020. The lifespans of the companies are:

• less than 1 year: 11 companies at 2.2%
• 1 to 3 years: 42 companies at 8.4%
• 3 to 5 years: 26 companies at 5.2%
• 5 to 7 years: 8 companies at 1.6%
• 7 to 10 years: 14 companies at 2.8%
• 10 to 15 years: 4 companies at 0.8%
• 15 to 20 years: 2 companies at 0.4%
• 20 to 30 years: none
• more than 30 years: 4 companies at 0.8%
• no data: 389 companies at 77.8%

What were the lifespans of the companies?

The data showed that majority of the companies shuttered within 1-3 years and 3-5 years. Many companies did not share their operating dates, or either their formation or shutdown dates were missing. Hence, 77.8% have no data.

What were the last funding statuses of the companies?

From our classification of data, the funding statuses are:

• seed: 107 companies at 21.4%
• early stage: 63 companies at 12.6%
• late stage: 17 companies at 3.4%
• initial coin offering: 1 companies at 0.2%
• M&A: 83 companies at 16.6%
• initial public offering: 3 companies at 0.6%
• private equity: 5 companies at 1%
• no data: 221 companies at 44.2%

The data revealed that many companies shuttered after they secured their seed, early stage, and M&A (merger & acquisition) financing rounds. Unfortunately, many companies did not share their investment information. Therefore, we were not able to verify the number of companies that did not receive any investment due to failed fundraising at seed stage, and those who struggled with lack of funds at early stage.

What were the total funding amounts raised by the companies?

Overall, a total of USD $11.8 billion (currency converted into USD) were invested in the companies:

• less than $1 million: 52 companies at 10.4%
• $1 million to $5 million: 56 companies at 11.2%
• $5 million to $10 million: 26 companies at 5.2%
• $10 million to $50 million: 68 companies at 13.6%
• $50 million to $100 million: 21 companies at 4.2%
• $100 million to $500 million: 19 companies at 3.8%
• $500 million to $1 billion: 2 companies at 0.4%
• $1 billion to $5 billion: 2 companies at 0.4%
• more than $5 billion: none
• no data: 254 companies at 50.8%

From the funding amounts raised by the companies, the investment size can be associated to the seed, early stage, and M&A (merger & acquisition) financing rounds in the venture capital market. Nevertheless, USD $11.8 billion were invested in 246 out of the 500 companies, and were forever lost.

How many employees did the companies hire?

From our classification of data, the number of employees hired are:

• 1-10 employees: 131 companies at 26.2%
• 10-50 employees: 141 companies at 28.2%
• 50-100 employees: 45 companies at 9%
• 100-250 employees: 37 companies at 7.4%
• 250-500 employees: 25 companies at 5%
• 500-1,000 employees: 14 companies at 2.8%
• 1,000-5,000 employees: 7 companies at 1.4%
• 5,000-10,000 employees: 2 companies at 0.4%
• more than 10,000 employees: 6 companies at 1.2%
• no data: 92 companies at 18.4%

From the number of employees hired by the companies, the data indicated that majority of the companies were startups or micro-enterprises (1-10), small enterprises (10-50), and SMEs (lesser than 250).

What is the gender statistics among the companies' founders?

From our classification of gender, we gathered a total of 938 data objects:

• male founders: 821 males at 87.5%
• female founders: 82 females at 8.7%
• founding corporations: 11 corporations at 1.2%
• no data: 24 at 2.6%

The data clearly showed that majority of companies were co-founded by male founders, particularly in the technology related industries.

Discussion

The findings of this descriptive analysis show that:

• companies with poorly designed business models have no product-market fit advantage and generally face strong competition
• as a result, the companies have a lower chance in securing investment funds and eventually shut down due to lack of funds
• shutdown rate is high among the startups, small companies, and SMEs during the early stage and post-M&A stage
• non-innovative new and small companies generally shut down within 1-3 years and 3-5 years of operation
• founders are predominantly male (87.5%) in co-founding a technology related business

Furthermore, the statistics seems to suggest that many new and small companies failed to identify their problem-market fit (aka problem-solution fit) before they could effectively validate their businesses to achieve product-market fit.

Problem-market fit is achieved when a company identified an existing real-world problem (critical pain points suffered by its customer segment) that is solved by its solution offering through deliberate processes of experimentation and validation.

Now, the interpretation of the reasons for shutdown is not straightforward because we merely observed 500 inactive companies. However, it is safe to say that most of the new and small companies often fail to design a good business model, and more often than not, there is no market need for their product-or-solution offering.

In fact, Marc Andreessen explained that "you can always feel when product-market fit isn't happening. The customers aren't getting value out of the product, word of mouth isn't spreading, usage isn't growing that fast, press reviews are kind of 'blah', the sales cycle takes too long, and lots of deals never close" (Griffin, 2017).

As a result, many of the companies from our database failed to achieve the elusive product-market fit, even though the market need was there and the product was compelling, but they just could not reach their customers (Feinleib, 2012).

That said, a majority of the reasons for shutdown are technical shortcomings which could be avoided through experiential learning, coaching, and mentoring. Therefore, this report implicates that failure may be part of the process of building a successful company, especially an innovative one.

Moreover, Triebel et. al. (2014) remarked that "every company founder should be aware of the fact that an important factor for establishing a company is the fault tolerance". They further explained that "fault tolerance, in the context of company foundation, generally describes a process of recognising, accepting, and learning from errors" (Triebel et. al., 2014).

In our future research, we will further investigate the effects of failure as part of the process of building a successful and innovative business. Especially when "failure must be accepted as a secondary effect" to founding successful companies, as advocated by Triebel et. al. (2014).

Our Conclusion

A descriptive analysis of 500 inactive companies have been conducted. Based on this analysis, we have shared our findings which highlighted a majority of the reasons for shutdown are technical shortcomings that can be avoided.

This analysis was limited by our small and incomplete datasets. With bigger and more complete datasets, we will be able to include inferential analysis, and eventually predictive analysis, to discover more meaningful insights.

Therefore, this report is presented as a work in progress. Nevertheless, our findings presented will be thoroughly expanded as part of our ongoing in-depth research.

As our data grows, we will continue to observe and examine the datasets to gain a better understanding on why companies around the world shut down.

Authors

1. Paul Lee, Co-founder & CEO, roiquant
2. Elina Kamaluddin, Business Research Analyst, roiquant
3. Sharifah Nadzirah, Business Research Analyst, roiquant
4. D. T. C. Lai, Co-founder & Chief Researcher, roiquant

Citation and republishing

Flipidea's articles and reports may be republished in accordance with our Citation Policy and Terms of Service.

References

If you would like to stay informed with our latest updates, subscribe to our newsletters or follow us on our social media (LinkedIn, Twitter, Facebook & Instagram).

This descriptive analysis report has been republished and featured as one of Hacker Noon's top stories on 24 June 2020.

Hacker Noon is the world's only independent tech blog with over 4 million monthly readers and average more than 8 million pageviews per month. According to Alexa, Hacker Noon is ranked in the top 3,000 websites worldwide.

Disclaimer: Any information, data and content on our analytics and business intelligence (ABI) platform, web applications or websites are for general information use only. The information and analyses presented on our web applications and websites do not constitute any legal, business, investment or tax advice. Even though certain information are cited from third-party sources while believed to be reliable, roiquant has not independently verified such information and makes no representations about the accuracy of the information or its appropriateness for any given situation. References to any securities or charts or graphs and all materials provided in connection with roiquant’s web applications or websites are provided strictly on “AS IS” basis without any representations or warranties, express or implied, which should not be relied upon when making any investment decision. Any projections, estimates, forecasts, targets, recommendations, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others.

Our proprietary data retrieval systems scan the Internet to automatically aggregate public information from news articles, websites, press releases, regulatory filings, and so on. Roiquant's intelligent decision support system, multimodality systems, database systems, data retrieval systems, algorithmic systems, data processing engines, and many other technologies are powered by artificial intelligence and advanced analytics using our own proprietary data and synergized public data. Roiquant does not warrant the completeness or accuracy of data provided, so you would need to independently verify the information. Nevertheless, we encourage you to check the accuracy of our information before its use. We also gently advise that you obtain sufficient knowledge, market understanding, professional advice and experience to make your own evaluation of the merits and risks of any actions with the information.

Last edited on 21 October 2022.