Skillr builds enduring teams ready for the future of work. This is our story.
In 2013, Megha and I started Leap Skills because we truly believed that opportunity and success should not be a function of where one was born. Over the past 7 years, we have acted on that belief in a tremendous way. Leap worked closely with a staggering 50,000 young jobseekers in tier-2 and 3 cities by providing them with the skills they needed to be successful at the workplace.
In an industry largely occupied by legacy players with classroom-only models, Leap was the first to leverage technology to scale interventions by developing a proprietary learning app that allowed jobseekers to interact with our programmes digitally. It was when we received the ‘MIT Inclusive Innovation Award’ for ingenious use of technology in skilling that we first began imagining a tech-led future for Leap, rather than a tech-enabled one.
Not only has Leap shifted to offering technical solutions to training providers today, but it also went on to raise over $2m from coveted institutional investors such as elea Foundation, Artha, Menterra, and NDSC. Appreciative that we were leveraging technology to solve India’s skill gap, tech veteran Caesar Sengupta, VP Product at Google joined us as an angel investor and advisor.
There’s not much I could have asked for at that point — we had a raft of investors by our side — a huge vote of confidence in the work we had set out to accomplish. And amidst the fundraising, targets, and firefighting every so often — my personal learning curve was off the charts too. Ostensibly, nothing needed change — if we continued to grow and work with the same gusto, things would be alright. But I started feeling differently after my time at GSBI.
GSBI (Global Social Benefit Institute) invites a select group of impact-driven startups to Santa Clara each year for intensive mentoring workshops to improve processes, drive value for beneficiaries, and scale operations. It is here when I first began to delve deep into a problem that had been lurking at Leap for a while: companies were getting harder to convince to recruit from Tier-2 and 3 cities. Even though trained jobseekers from smaller cities were equally competent, recruiters at these organisations were only interested in hiring from ‘good’ colleges. They were interested in a candidate’s credentials, over their potential.
Several conversations later, Megha and I concluded that skilling wasn’t going to be enough to address India’s skill gap. In order to ensure that every young jobseeker gets a fair chance at building a promising future, we had to approach the solution from another perspective: the employer’s. And that’s when Skillr came to life. Through Skillr, we aimed to reverse engineer the hiring problem from the customer’s perspective, by understanding their needs first. To be honest — we were initially surprised at just how consistent the employer problems were. They simply wanted talent that was:
- Reliable
- Trainable, and with
- Growth Potential
If we could quantifiably demonstrate that young jobseekers from across India — including those from non-brand name colleges and from tier-2 and 3 cities possessed these qualities too- then we would open a whole new stream of potential candidates for employers to choose from, while providing access to opportunity to the under-resourced. And so began Skillr’s journey.
The Problem. Our Solution.
Equipped with a team and resources dedicated to solve the employer challenge, we quickly realised that the challenge of finding reliable, trainable, and high potential talent — while still pressing — was felt most acutely with a certain segment of the workforce. The young workforce: millennials, Gen-Z and younger. And we discovered that the problem wasn’t merely as simple as ‘finding’ talent — engagement and retention were equally crucial.
Here’s some data to back that up. Employers on average invest 1.4x the annual salary on early-mid career employees each year, yet:
- 55% of young employees today are disengaged at work
- 60% of the young workforce today is open to switch jobs
- 66% of young professionals switch employers in less than 2 years
To put it another way — that’s an additional 40% invested in each employee to engage and retain them — but to no avail. This workforce is more disengaged than ever before — and much more likely to switch jobs without hesitation. And when you consider larger trends, this crisis of disengagement gets dilated: By 2025, 2 out of 3 employees in the workforce will be millennial or younger. So how can organisations address this issue? Incumbent talent management solutions just weren’t built to account for disengaged workforces and shifting preferences of employees.
Skillr’s Talent Intelligence Platform is built to address the acute pain and helplessness that employers experience when managing and retaining teams with young professionals. By leveraging decades of research in I/O psychology and our proprietary future-skills framework, Skillr is able to make informed talent decisions that help engage and retain young teams, including:
- Talent Identification: Hire talent that is quantifiably proven to perform in your organisation by predicting their skills in advance, or gauge the skill levels of existing teams. Identification of skills through assessments is the first step to enable a range of informed decisions on your talent.
- Talent Mapping: Quantifiable future-skills insights help map new and existing employees to job roles that leverage their strengths, while curating teams with diverse skill-sets in a way that maximises team potential.
- Talent Engagement: Once accurately mapped to teams with amplified potential, our platform interacts with your people and teams longitudinally to ensure ongoing feedback, timely recognition of people, and more meaningful engagement of your people overall.
- Talent Growth: Identify and nurture emerging leaders by tracking longitudinal performance of individuals and leveraging powerful insights needed to prepare and retain high-performing talent for the long-term.
7 years of dedicated research has led us to believe that investing in your talent through this 4-step process helps retain talent in the long-term — especially the younger workforce of today. Investing in talent through in an informed manner also has far-reaching financial benefits:
- Lower costs of hiring: Skills-based assessments can identify skilled talent, before you hire them, so you only hire talent that is guaranteed to perform. Future skills-based assessments predict performance more accurately than the domain skills-based ones that are traditionally relied upon. By filtering out candidates based on predicted performance, Skillr reduces recruiter and executive bandwidth spent on hiring talent.
- Lower costs of attrition: Skillr curbs the avoidable costs associated with attrition, such as the costs to re-hire and re-train candidates, and those associated with the loss of productivity.
- Lower supplementary costs: Reduce investment in costly external consultants to conduct processes such as appraisals, that can be better and more objectively conducted by Skillr.
Has this been solved before? Well, not really.
There are three different types of solutions attempting to address the problem of hiring, or retention, or in some cases both, but none holistically. In addition, none of these solutions were designed specifically to address the challenges associated with managing and retaining young talent:
- Agencies
- Hiring focus, so retention is beyond their scope
- Volume-based business model: They must deliver a certain number of candidate resumés to their clients, and often sacrifice candidate quality in the process
- The candidate selection process is inherently biased and unstandardised
On the other hand, hiring with Skillr promotes long-term retention. Our skills-based assessments objectively measure future skill levels to predict candidates likely to be successful long-term.
2. Assessment companies
- Primarily hiring-focussed
- They gather data on employees at sporadic intervals, so they unable to improve the accuracy of their hiring predictions over time
- Since they are used primarily for hiring, they are unable to provide managers objective talent strategies or assist with downstream decisions such as engagement and growth
Skillr’s longitudinal approach to engagement and data collection ensures that our algorithms are consistently adapting to improve the accuracy and quality of an organisation’s hiring over time.
3. Other Talent Intelligence Platforms
- Lifecycle focussed
- Their algorithms search for proxies within self-declared information, which itself can be inaccurate or misleading
- Another challenge of utilising self-declared information is that candidates are not being evaluated on their potential to perform, but rather their credentials on paper
Skillr leverages decades of research in I/O psychology and our proprietary skills framework to objectively measure the skills and merit of candidates, and not deductions based on self-declared information.
How we got here.
Since we began Skillr, we have rigorously developed and tested a range of algorithms underlying the Talent Intelligence Platform for their accuracy and predictability of long-term performance.
2019
In the early stages of development, we experimented with algorithms that scanned demographic data for ‘adjacencies’ or ‘proxies’, but after testing this with over 14,000 young job seekers, we had conclusive evidence to demonstrate that this method of approximating potential was inherently biased. For example: The algorithm began to suggest that jobseekers whose parents have been to college are automatically more capable than others. And while that may be true in some cases, this is definitely not an all-encompassing truth that can be validated by science.
We then began experimenting with the Big Five Inventory (BFI) framework created by psychologists at Berkeley. BFI is a psychometric framework designed to measure the Big Five dimensions (Openness, Conscientiousness, Agreeableness, Extraversion, and Neuroticism), through a series of self-reported responses. Widely acknowledged as the industry standard, we too initially believed that the results on the BFI could accurately predict performance at work. After testing this with over 4,500 young candidates, we realised that this model had its own limitations:
- While the model helped to understand someone’s personality better, performance predictions were often unwarranted extrapolations of data. Eg: Assuming that everyone who has scored highly on extraversion is automatically a good leader, manager, or even salesperson. It failed to draw compelling causal explanations between personality and human behavior on-the-job.
- These assessments are easier to decode by candidates, resulting in inaccurate personality assessments from the self-reported data.
- Psychometric assessments can themselves augment decision making, but cannot alone be left to predict on-the-job performance.
2020
We needed a framework that could accurately measure the skills and performance of people on the job, but psychometric frameworks like the BFI were not the answer. The framework also needed to be representative of the skills required for the future of work. So began our extensive research — we scoured through a range of different frameworks of 21st century skills — but none that we could directly apply to an Indian context. So we built our own: Skillr Instincts, a proprietary framework of 9 future-ready skills that collectively, and holistically could measure and predict performance at the workplace. We built the framework in conjunction with professors at Stanford and UPenn, who worked with us to reduce biases that can inadvertently creep into algorithms. And then begun our testing:
- We onboarded and monetised 30 employers who were looking to hire high quality, young talent, collecting 100,000+ data points in the process.
- Our skills demonstrated a composite reliability and internal consistency of 0.73, demonstrating promising early results that our framework can measure employee skills, but also that there is still a tremendous amount of work to be done in this direction
- The skills were also validated to measure a model fit using Confirmatory Factor Analysis in order to test how well the measured skills represent the performance indicator. The Comparative Fit Index (CFI) was found to be 0.80. This is a strong score and indicates that the nine skills in our framework collectively and holistically encapsulate the skills needed to succeed at the workplace. Thus, even measuring combinations of these skills that vary with job function can reliably reflect success on the job.
This testing helped iterate and improve the predictability of our algorithm over time. With compelling numbers backing the reliability of our framework, we were ready for launch.
Today
We then began a private beta with select clients, and demonstrated radical accuracy on two fronts:
- Objective People Decisions
- 95% accuracy in predicting the likelihood that an employee receives sales incentives. Going forward, this client can predict which candidates are likely to be successful in sales roles, before they are hired.
- 99% accuracy in predicting which employees were likely to get appraisals. Since appraisals essentially translate to high performance, this assessment gives our client the foresight to predict high performers in their teams, whether it be before hiring, for succession planning, or even more.
2. Cost Reduction
- 49% reduction on hiring costs alone by filtering out candidates who were likely to default, thereby trimming the unnecessary additional investment of time and money spent interviewing candidates.
Today, our early access programme is live and is available for select organisations to try free for 3 months, with promising initial traction:
- 4 companies have completed
- 8 companies under trials
- 22 companies in the pipeline
- Your company next >
So what’s in store for the future? For starters, our research team is now working to discover different ways in which the findings can be utilised to optimise employee development and interventions. In addition, the benchmarking of these findings will be used to extend the psychometric properties of our assessments in order to diversify the output to a more heterogeneous population, which we have had limited access to thus far.
To learn about our future updates real-time, subscribe to our mailing list. Questions about our algorithm? Our focus on the young workforce? Our data? Anything else? Let me know in the comments below 👇🏽
Thanks for reading!
Ankit