Predicting Startup Success with Data

, by Carlo Mammola and Luigi Mastromauro - Dept. of Management (Mammola); Accenture Italia (Mastromauro)
Some venture capital investors are starting to use data analytics to predict survival or failure. The authors have developed a data matrix identifying four startup archetypes

According to the latest research, 90% of startups fail and even those funded by venture capital funds (VCs) fail in 50% of cases.

Researchers have always sought to identify the factors that contribute the most to determining future business performance. In recent years, they have shown a growing interest in the world of startups, due to the increasing importance of venture capital.
By reviewing investors' decision-making process during investment selection phases, researchers have identified two main factors: economic performance (the business) and human capital (the team).

In a market where competition decides which companies are able to survive, the race to success requires a good horse (the business) and a good jockey (the team).
If the team is not well trained or the individuals within the team do not know how to work together, the startup will have little chance of overcoming the hurdles of its very first years. Likewise, if the product does not meet consumer needs or the entry strategy is poorly implemented, the chances of surviving decrease.

Given their close relationship, an intense debate has begun regarding which of the two factors is most important. Venture capital investors attribute the team's abilities as the main key to the success (or failure) of their projects, but not everyone thinks the same way. A famous investor, Warren Buffet, said: "When a management team with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact."

However, it should be emphasized that the analysis methods used in this sector are still traditional. "Data is the new oil" has become a mantra, yet VCs find it hard to use data to identify the most promising startups.
Even today, the process is based above all on networking and the experience of VCs.
Some operators, however, are beginning to apply data analytics to predict the future of startups. Their focus is to identify future business valuations or generate a score representing the probability of success.

On closer inspection, it is clear that there are many events that can influence future chances of success and that valuation methods are greatly weakened by the volatility typical of the first years of a company's life. So, how much does the aforementioned score help investors achieve the desired return?
In our analysis, the problem is approached from a different perspective: if there are in fact at least two factors that determine the success of a startup, a single score does not allow investors to properly analyze the team and the business.

The value of a bilateral approach is twofold: on the one hand, the pool of possible investments expands, as some of those traditionally considered to be of little interest can now be better analyzed. On the other hand, the risk of investment failure decreases, as it is immediately possible to identify the most problematic area in which to intervene, if necessary.

Therefore, to support VCs in their investment decisions, a matrix has been developed that identifies four startup archetypes based on their chances of success. "Premature" companies have low scores due to both the size of the team and business model performance. "Ethereal" companies are characterized by strong teams, but have poorly defined business models or products that are still incomplete. "Scalable" companies, on the other hand, show good economic performance, but have a team that is not suitable for sustaining their future growth. Finally, "Stars" are the startups everyone would like to invest in.

The challenge that awaits us, to finally make this a 2.0 sector, is to identify the variables that can quantitatively describe these two dimensions. This will allow venture capitalists to leverage their decisions against not only their instincts, but also a well-structured set of data.

Read more about this topic:
Nicolai Foss. Entrepreneurship Goes beyond Startups
Good Business Plan not Enough for Success. Interview to Alisée de Tonnac
Creating an Identikit of Entrepreneurs from 1850 to Present
Grow Your Business but Minimize the Risks
The Key Role of Taxation Regimes for Startups
A Small Fee for Business Training Improves Course Attendance
Family Ownership Helps Cope with Political Uncertainty
Experimental DecisionMaking Approach Attracts Funding
Entrepreneurs Who Act Like Scientists Get Better Results
Avocado Toast Leads to Hot Brooklyn Startup. Interview to Alessandro Biggi
Turning "Spotify for Textbooks" into $4.8 mln. Interview to Perlego
Looking at Gender Bias in Funding Yields a Surprise Result
Family Firm Governance Results Linked to Context
Social Networks Foster Entrepreneurship
Taking Texting to the Next Level with Kaleyra. Interview to Dario Calogero
Connecting the Dots on How WWI Unleashed new Business Performance