Can You Predict Which Startups Will Make It?

René Bohnsack thinks so.

He’s designed an online platform that combines crowd-sourced feedback, expert consulting, and predictive technology to determine which business models are most likely to succeed. By shuffling the elements of the model—from partnerships to customers, key activities to resources required—entrepreneurs can place different bets and apply the winning combinations to their startups. 

Bohnsack, an assistant professor of strategy and innovation at Católica-Lisbon in Portugal and an International Faculty Fellow atMIT Sloan, has found in his research that venture capitalists often rely on assumptions and cognitive biases rather than data to decide which ideas to fund. As a result, they can overlook good ideas and entrepreneurs can fall into a trap of developing businesses that play to investors’ assumptions—but aren’t necessarily destined to succeed.

As part of the IFF program, which invites faculty members from MIT Sloan's partner institutions to Cambridge for a semester of teaching and research development, Bohnsack has worked closely with the MIT Real Estate Innovation Lab to build out the machine learning aspect of his platform. He has also integrated more parameters for entrepreneurs to experiment with as they fine-tune their models.

With the open, free, web-based platform, Bohnsack aspires to help good ideas anywhere in the world come to life and to create economic value and opportunity as they grow into successful companies. With his focus on ideas and impact, he says, “My time at MIT Sloan is serendipitous. It’s like coming home.”

IDEAS MADE TO MATTER