Culture

Coding Inclusion: An Interview with Develop Diverse Founder Jenifer Clausell-Tormos

Catier Women’s Initiative is a yearly program funded by French luxury brand Cartier to support women in entrepreneurship, with a focus on women-owned and/or women-led business that drives social change.

The annual competition includes businesses from all over the world. “Issue areas” include: poverty, hunger, health, gender equality, clean water, clean energy, climate action, responsible consumption, innovation, and much more.

One of this year’s finalists is 38 year old Spanish-born, Denmark-based Jenifer Clausell-Tormos, the founder and CEO of Develop Diverse.

Develop Diverse uses the latest in machine learning to help companies unravel unconscious bias in their language and content, particularly around hiring practices.

The result is that companies are able to employ more inclusive language and attract a more diverse workforce in terms of race, gender, neurodiversity, age, and physical ability.

Find out more about Develop Diverse and what it means to be part of the Cartier Women’s Initiative:

Tell us about your work and education background.

My name is Jenifer, I am the CEO and Founder of Develop Diverse.

I studied Chemistry in Valencia, Spain, and moved to Strasbourg, France, where I did a PhD. My focus was developing tech platforms for biomedical applications to accelerate the process of finding new medicines for different type of diseases. I worked in this field for 10 years in Madrid, London, and Copenhagen.

Four years ago, while working and living in Copenhagen in biotech, I decided to use my expertise in tech to accelerate the process of closing the gender and diversity gaps in organization.

That brought me to found Develop Diverse.

I speak Spanish, Catalan, French, English, and Danish. I am a TEDx speaker, am part of the Thinkers50Radar 2021, and was nominated to EY Social Entrepreneurship award in 2020.

 

 

 
 

When and why did you launch Develop Diverse?

I started Develop Diverse in 2017 after realizing that gender roles were still prominent in my generation. I knew that this was a thing in my mom’s generation but not in mine, where women having a higher education and economic independence is more common than ever before.

I wanted to help every single person unlearn their unconscious biases, so we can break down stereotypes in just a few decades rather than wait another hundred years.

Unconscious bias trainings were the method most used to do that, and clearly it was not enough given that the diversity numbers had not changed much for the last 100 years. I decided to use tech to accelerate the process. In my career, I always used tech to accelerate the process of finding new medicines for specific diseases, so why not to use tech to close the diversity gap?

Peer-reviewed research had shown that we write job descriptions using gender-specific words, and these words, not the skills required but the words only, impact who is going to apply for a job.

I thought ensuring inclusive language could fix this because language is powerful and therefore, having inclusive language would allow us to build a culture of inclusion in the workplace, and eventually, an inclusive world.

 
 
 

Can you share a bit about the development process? How long did it take to get right, and how many iterations were there before you felt ready to launch?

In general, we believe development never stops to ensure our customers’ needs are always covered. If you’re referring to the initial development, it took us about two years.

We launched the software at the end of 2019. During this time, we developed the strong engine of our software. We had many iterations with our early adopters that allowed us to develop the right design and the most relevant features.

The biggest milestone was proving the efficacy of our software with our early adopters. One of them was Amazon Europe. They tested our software in 2019 in the UK and Germany to see if they could get more qualified female applicants in management- and STEM-like positions.

The results spoke for themselves. They got almost 4x more qualified women and 1.5x more qualified men of diverse backgrounds. They managed to close the gender gap in the set of job openings in which they tested the software.

 

 
 

What are some of the keywords that Develop Diverse looks out for, and how do you think changing them makes a difference within these work places?

In job descriptions, biased words prevent us from attracting diverse talents. Some examples of these words are: “Competitive company” or “ambitious teams.” They are more attractive to men while they discourage women. If, however, we use inclusive terms such as “aspiring company” or “motivated teams,” our job ads will be appealing to all genders.

Why is this happening? Traditionally, men are associated with agentic traits, achievement, and individualism. While women have been associated with communal traits, and are expected to take care of the family and ensure harmony.

This affects how people perceive words based on the gender they identify with.

It is important to highlight that this doesn’t mean that women are not ambitious, or men are not nurturing – it only means that due to gender stereotypes, they do not feel represented by certain words.

Develop Diverse aims to help people make an informed decisions about what words are biased so that they can change them if their goal is to increase diversity and inclusion.

Our customers, when using Develop Diverse for their job descriptions, experience an increase of diverse qualified applicants almost from day one.

Language is a powerful tool, so if we want to build a culture of inclusion we need to ensure that we use inclusive language in our communication.

I would like to add that Develop Diverse goes beyond gender. It also helps closing the gap for other diversity groups, such as age, ethnicity, non-conforming gender, physical disabilities, and neurodiversity.

 

 
 
 

How do you ensure the Develop Diverse algorithm is, itself, not biased?

It is a legitimate concern that language models reproduce and amplify biases existing in data. This concern is all the more valid the bigger the dataset in question is.

In fact, we leverage the biases that can be found in existing models or datasets and use it for detecting biased phrases.

At the same time, we make sure that our own models do not reproduce any of these biases by correcting them with our science-based inclusive language database that is curated by our linguistic team. As such, the only ways our tool changes the original texts are limited to inclusive language.

 

 
 

What happens if a company institutes Develop Diverse but sees no change in their applicant pool? What would be your next step with a situation like that?

This is not something we have reason to believe will happen, but we work with all customers to make sure they reach their goals and work with the software and recommendations given by the software the best way possible.

In a scenario where the applicant pool would be the same after implementing Develop Diverse, we would dive into the analysis alongside our customer to understand the story told by the data and find places to optimize further for measurable results.

 
 
 

Do you approach companies, or do they approach you, to use Develop Diverse?

It is a combination of both. At the moment, we mostly approach companies.

Companies are aware of unconscious bias being the bottleneck to getting diverse talent, but they may not know that there is a tangible solution.

We encourage companies reading this article to write to [email protected] and we will be able to explain to them how easy is to change the numbers when using Develop Diverse.

 

 
 

What kinds of companies are you seeing use Develop Diverse, and what kinds of companies do you hope will use Develop Diverse?

Today we are working with some of the largest brands in Europe. We see a large demand from male-dominated industries and companies that rely heavily on recruiting from Science, Technology, Engineering and Mathematics (STEM) fields, as these companies struggle to attract a diverse pool of applicants and meet their internal diversity goals.

Some examples are global corporations like Amazon, Vestas, and Russell Reynolds. Other are global scale-ups, such as Trustpilot and Templafy. Some examples of Nordic corporations are Danske Bank, TDC, ATP, Energinet, and Aalborg University.

We believe that eventually companies within other industries that are less male-dominated will use Develop Diverse, since diversity is more than gender diversity. It also includes age, ethnicity, sexual orientation, physical disabilities, neurodiversity, and more.

 

 
 
 

What are your goals with Develop Diverse?

Our firm aspiration is to grow fast and sustainably to achieve our vision of normalizing diversity with inclusive communication.

Today, we support businesses in improving their recruitment pipeline and talent retention by providing them with software that eliminates discouraging language in their external communication. This has been shown to dramatically increase the amount of qualified diverse candidates in all parts of their recruitment pipeline, and also bring awareness of unconscious bias among users.

In the short-term, our goal is that our software is used by every single employee, so companies can not only attract diverse talent but also build a culture of inclusion by ensuring their whole organization communicates inclusively.

We believe this will help organizations increase employee engagement and happiness, as well as increase companies’ market share since they will also communicate inclusively to their customers.

In the long-term, our goal is that our software is used by every single person, for any text type, for any communication channel. We want each person to communicate inclusively and become conscious of their own unconscious bias through the language they use daily. It will be like having a personalized unconscious bias trainer with you all the time.

 

 
 

Can you tell us a bit about your experience with the Cartier Women’s Initiative and what you hope results from this experience?

I feel honored and proud to have been nominated top three in Europe. Cartier Women’s Initiative is enabling us to showcase our work and goals globally. We also got lots of great feedback on our business from Cartier Women’s Initiative coaches and advisors.

In addition, I got access to a huge network of other interesting female entrepreneurs worldwide. We exchange ideas, share experiences, and support each other.

 
 
 

What advice would you give other women who may want to apply for this grant in future? What advice would you give other women interested in starting a tech company like Diverse Develop?

I would encourage other women, who have a start-up and find Cartier Women’s Initiative attractive, to apply. I would advise them to not question if their business is not advanced enough or good enough; they should just apply!

I postponed the application and was waiting for the moment it felt right. Luckily people in my network encouraged me to do it; this prevented me from missing this great opportunity. It turned out I was wrong that we were not ready; we were very ready!

My advice to encourage women to build a tech start-up would be to share the following knowledge:

Women being the minority in tech start-ups and in entrepreneurship overall is not because women are not capable. The reason is that women have not been encouraged to go into tech, and rather focus on other industries like health and education. This results in a lack of women role models in tech and other STEM fields.

It is important to note that in this statement I assumed binary gender, but I would like to acknowledge the existence of more than two genders in our society, and it is important that we are inclusive of all gender identities.

See our series Ladies Who Launch to learn about more women in business.