Data-centric Recruiting

In 2019, a prominent company in Iran approached me to overhaul their hiring process. Although they were one of the pioneers in their industry, they were struggling to attract and retain talent due to perceived biases and an outdated recruitment experience. My goal was to design a data-driven, bias-resistant hiring system that would make the process fair, inclusive, and appealing to a broader talent pool, ultimately restoring the company’s competitive edge.

Initial Challenges

The company's legacy hiring process had several critical issues:

• Lack of Diversity: 90% of employees were male, with no employees with disabilities.

• Employee Turnover: 20% annual attrition due to dissatisfaction with role fit and team dynamics.

• Failure to Meet Goals: Only 40% of employees met their position goals.

• Cultural Barriers: Iranian cultural attitudes toward disability and gender contributed to a lack of trust among women and candidates with disabilities.

These factors led to three core pain points:

1. Inability to Attract New Talent

2. Lack of Workforce Diversity

3. Retention Challenges with Existing Talent

Research and Process

Double Diamond Design Methodology

I implemented the Double Diamond approach to structure our process into four stages: Discover, Define, Develop, and Deliver.

Discover

Our research began with analyzing company data, industry benchmarks, and competitor employee demographics. To build a comprehensive view:

• Surveys were conducted with the hiring pool, revealing hesitations among women and people with disabilities due to perceived biases.

• Interviews and Ethnographic Research: We observed the company's hiring system firsthand and interviewed HR, hiring managers, and employees to understand internal perceptions.

Key Insights:

Our research began with analyzing company data, industry benchmarks, and competitor employee demographics. To build a comprehensive view:

• Bias Concerns: 70% of women felt gender bias limited their chances. 90% of potential candidates with disabilities avoided self-identification, perceiving it as a risk to job offers.

• Demographic Imbalance: The company’s workforce was 90% male, with no representation of people with disabilities, which was contributing to a stagnant and less innovative environment.

Define

With these insights, we defined the primary problems:

• Bias in Hiring Process: Existing steps allowed for implicit biases based on gender, and disability.

• Outdated Assessment Criteria: Candidates were evaluated on personal demographics before skills, which limited diversity and overlooked talent.

I crafted a new hiring process designed to prioritize qualifications and skill assessments over personal demographics, mitigating bias and improving trust.

New Hiring Workflow:

• Initial Application: Candidates submit only education and job history.

• Personality and Skills Testing: Objective tests to assess role compatibility and skills.

• Demographic Information Deferred: Demographic details, including gender and race, are only disclosed after skill-based acceptance.

• Disability Accommodation Question: Instead of direct self-identification, all candidates are asked if they can perform job functions with reasonable accommodations.

• Interview Stage: After demographic information is shared, qualified candidates proceed to interviews based on objective performance metrics.

Develop

After stakeholder alignment and multiple rounds of A/B testing, we developed a user interface that integrated seamlessly with the company’s design language. The new UX was rolled out in a 3-month pilot program.

Deliver

Results

The new system drove significant improvements during the test period:

• 75% of New Hires Were Women: Gender diversity increased dramatically, addressing previous gender imbalances.

• 25% of New Hires Were People with Disabilities: This helped to build an inclusive workforce and improve employee morale.

• Enhanced Retention and Satisfaction:

  - 80% of employees met their position goals after implementation.

  - Employee turnover dropped from 20% to 5%.

  - Role fit satisfaction improved, with only 20% of employees changing positions compared to 80% before the redesign.

The impact extended beyond initial expectations. The company regained its position as the market leader, with a newly diversified team contributing fresh perspectives and innovation.

Outcome (2023 Snapshot)

Today, the company has grown from 500 to over 1,600 employees:

• 50% of employees are women (up from 10%).

• 10% of employees are people with disabilities (up from 0%).

• Sustained Retention: The majority of employees remain in their roles, indicating improved satisfaction and role alignment.

What We Delivered

1. Data-Centric Recruiting System: Eliminated biases against race, gender, nationality, and disability.

2. Skill- and Qualification-Based Hiring: Emphasized objective testing over demographic indicators.

3. Inclusive Accommodation Practices: Replaced disability self-identification with an accommodation question, fostering trust and inclusivity.

Reflections and Takeaways

In working on this project, one challenge was managing cultural resistance and advocating for inclusivity in an environment with limited protections for women and people with disabilities. In hindsight, I would push for more time to allow for deeper candidate-focused UX research. While the project was a success, balancing advocacy with tight timelines required extensive stakeholder management, which affected our ability to fully optimize the end-user experience.