STUDY | Sharpening the selection of innovation projects can bring unexpected setbacks

Choosing which innovation projects to support is one of the most important – and difficult – tasks of accelerators. It is not only about finding the winners, but also about avoiding the costly mistakes.

In one study, based on 3 580 applications and 126 interviews, researchers followed an accelerator that had been trying to improve its selection process for several years. They introduced more assessment steps, put more emphasis on the contractor’s credentials and built in a pre-screening of applications. The goal was clear: higher quality applications and fewer wrong decisions. And yes – the applications got better. But the decisions did not.

Worse assessments with stricter selection

One of the main findings of the study is that the most rigorous selection model – the one with the highest thresholds, toughest screening and greatest focus on track record – led to more wrong decisions, not fewer. The researchers identify two explanations:

1. overconfidence in past performance
Assessors began to lean heavily on the entrepreneur’s past successes. The problem? Early successes turned out to be partly random. This meant that promising newcomers were weeded out, while some more established entrepreneurs were given too much credit.

2. A tougher system changes who applies
When the process became tougher, two things happened:

• Experienced teams submitted more half-hearted projects – using the accelerator for sidetracks rather than ventures they really believed in.
• New entrepreneurs with good ideas chose not to apply, believing they would not be able to compete with the ‘stars’.

The result was an application funnel where the quality of the paper increased – but where the decision-making evidence actually became more difficult to interpret.

Three lessons for accelerators

The study shows that selection processes are more difficult than you might think. Here are three recommendations from the study for accelerators looking to develop their selection process:

• Be careful not to overestimate past successes. They are not a guarantee of future results.
  
• Keep in mind that changes in the sample affect who applies. Too strict a filter can scare off promising newcomers, while experienced teams submit projects they have not fully invested in.
  
• Introduce changes gradually and monitor their effects closely. The accelerator in the study changed several things at once – assessment steps, criteria, screening. This made it difficult to see what effect each adjustment had.
  
  
  

More about the study and the authors
The article Selection Regimes and Selection Errors is published in the scientific journal Organization Science. The authors are Dmitry Sharapov, Imperial College Business School, UK and Linus Dahlander at ESMT Berlin, Germany.