The hard truth about startup failure in Europe
When you are involved in the startup world you might have heard someone claim that nine out of ten startups will fail. While the startup world always has agreed to this in general, there is no data to back this up. If you are a founder, an investor, a corporation or a company builder — as we are — you want to understand exactly what is the likelihood of success. And you also want to know how you can increase it. As a founder or a corporation, you might think that an accelerator program could be the answer.
We took the challenge and a closer look at the data!
For our research we primarily used Crunchbase as our source. Crunchbase is one of the largest startup databases available and has listed more than 900,000 companies (as of March 2020). The Crunchbase database generates its data from their investor network, which contains more than 3,500 investors, the community, AI tools and a data team.
We filtered the database for companies in the EU28 area that were founded in 2013 and divided the data into startups that were in an accelerator program and startups that did not go through an accelerator program. In order to get a significant sample, we randomly selected 400 accelerated and 400 non-accelerated startups. We decided to take the data set from 2013 because, depending on the source and geography of data, startups need five to seven years (median) to achieve an exit.
We subsequently classified the startups into the following groups using not only Crunchbase, but also adding data from LinkedIn and publicly available information to our analysis:
- Exits: are flagged as exited by Crunchbase
- Scale-ups: flagged as alive on Crunchbase and 50 or more full-time employees on LinkedIn
- Somewhat alive: flagged as alive on Crunchbase and between ten and 49 full-time employees on Linkedin
- Zombies: flagged as alive on Crunchbase and less than ten full-time employees on LinkedIn
- Dead: flagged as dead on Crunchbase or flagged as alive on Crunchbase, but no operational website or LinkedIn page and no full-time employees on LinkedIn
For analysis and interpretation of the data, we only consider Scale-ups and Exits as being successful. Of course, we know that the number of employees on LinkedIn is a simplified view to cluster startups in different categories and that there are startups that managed to get really big with a relatively low number of employees, Instagram for instance. But in general, it is a good indicator for traction or success of the startup in the European context.
True or not — do nine out of ten startups fail?
According to our data, 29% of the European startups founded in 2013 are “stone dead”. 38% can be classified as Zombies and 22% are Somewhat Alive. As per our definition, this amounts to an overall failure rate of 89% in our data set. On the success side, only 6% of startups founded in 2013 have made it to an exit. Another 5% are still in a Scale-up phase. These numbers add up to a success rate of 11%.
So, our data ultimately proves that “nine out of ten startups fail”, explicitly in the recent European context.
Do acceleration programs increase the odds?
For those who are not familiar with the term: accelerators are mostly institutions that support startups through intensive coaching over a certain period of time, thereby having the goal of greatly accelerating and driving the startup’s development process. Accelerators can help with both knowledge and resources. Depending on the set-up, accelerators take equity from the startup in exchange for participation in the program.
Getting accepted by a (top) accelerator usually means that a startup complies with the respective acceptance criteria, i.e several factors are taken into account during the admission process in order to pick the most promising ventures to take part in the program.
So, shouldn’t we assume that the chances that an accelerated startup will succeed are much higher than those of a non-accelerated one? First, the startup needs to be accepted by the accelerator, and secondly, it can benefit from additional resources in order to boost its overall development.
Our data do not support this assumption. In concrete numbers: the accelerated ventures outperform the non-accelerated ones by only 2% — which is neglectable. The results (table 2) show that there is basically no difference in success rate if a startup was accelerated or not.
Hence, as a founder it is worth asking the question: If there is no upside, why would you want to give away equity by joining an accelerator?
Corporate accelerated success
In order to have a better overview of the whole picture, we decided to analyze the startup success rate for corporate accelerators as well, i.e. splitting the data for accelerated startups into “corporate accelerated” and “non-corporate accelerated” startups. A corporate accelerator is a specific form of an accelerator, usually run by a for-profit corporation.
The biggest difference between a regular accelerator and the corporate one is the objective. While the first one acts more like an entrepreneurship bootcamp by offering startups guidance and helping with resources, the second (corporate accelerator) aims to give corporates access to upcoming technologies provided by the startup. In exchange, at least in theory, the corporate helps the startup by acting as and / or finding clients or distribution partners.
On that basis, we could expect the corporate accelerated startups to perform even better than those going through a regular accelerator or those not going through an accelerator at all.
However, the results from our analysis indicate: corporate accelerated startups tend to perform even worse compared to the other two groups of startups.
As our data set is limited in this case and is only sufficient to rather show a first indication than to give a final statement, we decided to have this indication challenged by an expert. One of our partners at Stryber — Garan Goodman — ran two corporate accelerators for several years before joining Stryber, which makes him the right expert to listen to in this case:
“The success of startups in corporate programs is dependent on two factors:
Firstly, the agreed objective of the accelerator: For instance, Telefonica Wayra was to focus only on financial return. One of the results was the last mile pioneer Foodora which quickly exited to Rocket Internet and scaled to a 20 country business. In the case of MetroTarget the objective was to run pilots with late stage startups that had proven product and services in new innovation and a reliable team. Working with eight late stage companies the results were six pilots within three months of the program finishing. This was only possible in the context of the accelerator which works closely with the MetroTarget CEOs and their teams. Normal purchasing departments would have struggled to come close to the same results. Secondly, a suitable accelerator leadership: Often the corporate will appoint a senior manager. Unless this person has authentic founder experience they will struggle to have the right network which leads to startup selection, investment, mentor and focus on her accelerator business.
Corporate acceleration can be highly successful but the right objectives and the right leadership must be in place, just like any authentic business.”
The hard truth
- Nine out of ten startups will fail (now finally confirmed!)
- Accelerators don’t have a significant impact on startup success
- There is something wrong with corporate accelerator programs, based on an early indication
With nine out of ten startups failing, all ecosystem stakeholders — founders, investors, corporations and company builders — just have to live with this darwinistic rule of thumb, which now has been backed by data. Especially, investors and company builders do have a powerful strategy in place to cope with it: they spread their risk over a portfolio of startups.
For founders, unfortunately, we cannot identify joining an accelerator program as being a relevant strategy to increase their odds, in general. Every founder team will have to weigh the pros and cons of the respective program of interest. For sure, giving away equity, which is a prerequisite for some programs, will be a very high price to pay.
Corporations have to realize that their value-add, for now, is highly questionable. The benefit cannot be one-sided. Their programs have to be designed for startup success, for which there are at least two more stakeholders to be considered: the founders and their investors. We strongly suggest that corporations review their accelerator programs. In the long run, there is no point of participating in them, if the odds are not significantly greater than with non-accelerated startups.
About the authors: Alexander Mahr, Betina Puka, Matthias Dörner, and Simon Sommer are colleagues at Stryber — a leading Corporate Venture Builder based in Germany, Switzerland and the Ukraine. Alexander Mahr is Co-Founder & Managing Partner, Betina Puka, Simon Sommer (Munich office) and Matthias Dörner (Zurich office) are Venture Architects with hands-on experience in launching startups in both a corporate and a stand-alone context.
