flying slippers blog

Blind Applications, Impartiality, and De-Biasing a Selection Process

February 6, 2020

Blind auditions for orchestras have widely been heralded as reducing the gender gap by reducing potential biases in those selecting the musicians. But what does it look like when the same approach is associated with a tech accelerator – when gender is just one of many demographics that could be leveled?

 

In September we introduced our newest startup accelerator, Launch Lane in partnership with some of Philadelphia heaviest hitters – representing Backstage Philly, BCK Group, Mogulette all while leveraging the Science Center’s successful Digital Health Accelerator platform.  

 

Following a multi-step process including a blind application, we’ll announce the companies selected for the first cohort of Launch Lane later this month. Before making that big announcement however, we wanted to give you an inside look at what an anonymized application process looks like and how it impacts demographics.  

 

The blind application process isn’t new to the Science Center. It’s been implemented in the three cohorts of our Digital Health Accelerator and has proven to be effective in identifying promising startups with diverse founders. In fact, over 50% of DHA graduates are women or minority-led and of the 20 that have gone through the program, 19 are still in business.  

Solicitation  

Before we get into the demographics of the applications, let’s take a look at our outreach plan to attract those applicants. We were resolute in that the “Field of Dreams” approach would not do: just because we built it, does not mean they will come. They come if they feel the accelerator represents them, represents their interests and is run by people they trust. This is why assembling the leadership we did was so crucial to the success of Launch Lane.  

 

Further, forged partnerships with Black Girl Ventures, Black Female Founders, Black and Brown Founders, Code Fever, and Lightship Capital and represented Launch Lane at a number of events around the country including AWS Day, African American Chamber of Commerce, i, WURD Founder’s Day, PA Governor’s Economic Empowerment Summit, UNCF Innovation Summit to name a few – this alongside all our traditional outreach channels. 

 

Last, we set up post at Venture Café every week for anyone interested in the program to drop by, ask questions and meet the team in a low-stakes environment. As such, we had informal meetings with roughly 100 potential applicants during the solicitation period.

An application process that strips founders of any of the identifiers may be the best way to ensure all the innovators reaching out to us get a fair shake at valuable resources to grow into thriving companies.

Science Center

The Selection Team 

While we were out promoting Launch Lane, we were also hard at work assembling a selection team that was diverse in race, gender, geography and expertise. The Launch Lane Selection Team will remain anonymous, but we can share that it’s made up of 31 people of which 77% belong to underrepresented groups.

 

The Process 

Applicants submitted an initial application that was stripped of any pronouns or identifying information. This means that the selection team didn’t know if candidates were male or female, attended their alma mater, were from the same hometown and so on. 

 

Once the selection team reviewed all the applications and narrowed it down to the second round of finalists, we asked each applicant to complete a demographics survey. This is how we collected all the information we share below.  

 

Decisions were made by reviewers without knowledge of the entrepreneurs they were evaluating (aka blindly) through the anonymized document they submitted and before receiving the demographic survey.

 

Evaluating our Process 

We’ve engaged a third-party evaluator to help us improve and refine our process in the future, just as the startups we serve are constantly iterating.

The Numbers

By all accounts the pool of candidates were impressive and we are fortunate to have such a strong selection to choose from.

 

We received 206 applications that were then narrowed down to 47 through the blind process.

 

Here’s a look at the breakdown from initial submissions to first cut.  

206 Initial Applicants 47 Semi-Finalists
GENDER
Male 71% 49%
Female 29% 51%
RACE
White 27% 33%
Black 34% 32%
Hispanic/Latino 8% 9%
Asian 17% 21%
Middle Eastern 6% 8%
EDUCATION
High School 6% 4%
Bachelors Degree 32% 28%
Masters Degree 32% 57%
Ph.D. or higher 9% 13%

There are some stats we’re particularly proud of in these initial results and some we know need to be improved upon. 

 

What’s next

The anonymized portion of the application process is now over. As the applicants are narrowed down to 15 finalists, they’ll be invited back for an additional round of interviews.   

 

An application process that strips founders of any of the identifiers may be the best way to ensure all the innovators reaching out to us get a fair shake at valuable resources to grow into thriving companies. In a few weeks, when the selection team chooses the class, demographic trends can be better analyzed. From there, our sole objective is to launch these startups into a growth trajectory that knows no limits.