BCG & pymetrics on Unlocking Human Potential with AI-Powered Talent Decisions

Jacquelyn Soh
June 24, 2021

The rise of automation has fueled a business frenzy to seek talent with the latest digital skills. However, future-facing organizations should also be looking out for soft skills in their candidates like the ability to learn and adapt quickly.


Amber Grewal, Global Head of Talent at BCG, offers this outlook on talent potential in a recent discussion at CogX. Speaking with pymetrics co-Founder and CEO, Dr. Frida Polli, Amber described why the global consulting powerhouse is fully embracing AI to make more effective, inclusive, and future-proofed talent decisions.


“At the macro level, not just BCG, we see this ongoing technology advancement and market transformations. It’s requiring flexibility in thinking, quick and continuous learning, and mobility.” Amber added, “Jobs themselves are constantly evolving, and skills are changing. The average shelf life of [hard] skills is four to five years…Research has shown that about 87% of companies are going to be experiencing skills gaps.” 

For employers, therefore, Amber urged that now is the time to start looking at soft skills to prepare for the changing landscape. 


A new people science focused on soft-skills


Soft skills are the cognitive, social, and emotional characteristics inherent in people that are indicators of behavior and potential to succeed in different roles.


Although 92% of talent professionals and hiring managers agree that soft skills are just as (if not more) important than hard skills, only a quarter are confident in their abilities to identify them in candidates and employees.


For BCG, this is the game-changing potential of pymetrics and the reason why they have piloted the platform in multiple markets.

Dr. Frida Polli founded pymetrics to fill this gap. Trained as a cognitive scientist at MIT and Harvard, she knew that measuring behavior can help to quantify soft skills in a way that was once reserved for technical abilities. The result is a more accurate, 3- dimensional view of a person’s potential that far surpasses what resumes can reveal.


For BCG, this is the game-changing potential of pymetrics and the reason why they have piloted the platform in multiple markets. pymetrics enables BCG to uphold a rigorous selection process, while expanding the funnel to include more diverse candidates.


Advancing diversity with BCG’s Bionic talent model


D&I is an important topic to address for BCG, and one close to Amber’s heart.


As an organization with a varied portfolio of talent around the world, it is vital that the company commits to “clear actions and transparency”. This starts with what Amber refers to as BCG’s “vision and strategy of a Bionic talent model”.


A Bionic talent model integrates key outcomes across the talent lifecycle. It connects high-value human solutions with technological innovations such as AI to make accurate, data-driven predictions about talent supply, skills shortage, sourcing strategies and so on.


Applying the bionic vision to recruitment, BCG has designed simulations to allow more candidates to understand the role of a management consultant, implemented predictive hiring solutions like pymetrics to broaden their paradigm of ‘high-fit’ talent, and adopted a range of other digital tools to create a more positive and personalized experience. Bionic has enabled the organization to attract, select, convert, hire, and develop their best and most diverse talent to date.


“BCG is not the same company it used to be 10 years ago,” Amber said. “We’ve expanded our digital footprint and our GAMMA business, how we are looking at cloud technology and AI. We ourselves have also moved into a Bionic model that [requires us to] converge our skills – both the digital and the consulting.”


Using AI for good, always


Whilst AI’s ability to unlock human potential is well-documented, so is the risk of perpetuating bias when it is improperly designed and implemented. The speakers tackle the concerns head on in the latter half of the session.


“Algorithms are trained on real world data”, Frida explained. “If data in the world is historically biased, it will then potentially lead to biased algorithms.”


On the bright side, there are principles that companies can abide by to mitigate the risks. Frida shares the three-step approach taken by pymetrics to enhance fairness in AI:

Step 1: Rely on data that is equally distributed across gender, race, and socio-economic status, i.e soft skills data.

Step 2: Optimize algorithms for both performance and lack of bias; do not release an algorithm if it does not meet statistical parity, or group fairness standards.

Step 3: Monitor the algorithms once they are deployed to ensure that they continue to show group fairness.

The objectivity and fairness of pymetrics’ algorithms was recently affirmed in a third-party audit by AI experts from Northeastern University 


Adding a corporate lens, Amber shared her learnings from incorporating AI into the talent strategy of not just BCG, but also IBM and Intel where she previously led Talent Acquisition. 

For her, it starts with being clear about the outcome you want to achieve. Then, identify the data sources that support this outcome and evaluate the quality of the data. In the case of recruitment for example, your ATS can only provide you with historical data, whereas assessments might add the real-time layer necessary for more accurate decisions. Finally, Amber cautions that data privacy must be kept top of mind throughout the process. 

Creating human-centered talent experiences with AI

Contrary to skepticisms, AI can in fact result in more meaningful interactions throughout the talent journey at a scale that was hitherto impossible. 

As evidence, Frida pointed to how pymetrics is able to provide personalized feedback to every single candidate and employee who completes the games. Furthermore, the platform can redirect individuals to better-fit opportunities if they turn out to not be a match for the original role.

Amber reinforced the argument by sharing how BCG has deployed AI that can provide tailored coaching to applicants through the hiring process. For instance, the company rolled-out chatbots that could assist candidates in preparing for the assessments and interviews, and even recommend BCG-ers whom they could follow or network with. 

She asserted, “There is so much value-add. We are seeing it from a diversity perspective, we are seeing it from a quality perspective, from a speed perspective… [Automation] is giving our teams the opportunity to do more high value things like talking to the candidates and talking to the business and solving all those problems.”

“At the end of the day it’s about people. It’s about talent. They are the potential that every company is looking for.”

(For a bonus segment on what Amber looks out for in a tech partner to ensure successful implementation, watch the session here.)