Research Snapshot: AI-Driven Skills Taxonomy Checklist: An Evidence-Based Guide

A well-defined skills taxonomy, or framework, helps organizations standardize skill definitions, align talent with opportunities and drive workforce transformation. Developing a skills taxonomy using generative Al tools like SkillsGPT or WorkforceGPT requires a structured, evidence-based approach for effective implementation. Here are best practices to support your organization's skills transformation journey:

Begin with a Proof of Concept (POC) or Pilot

Test Al's ability to assess and map skills by focusing on a core skillset, allowing for adjustments before expanding to additional skill categories.

Example: In Johnson & Johnson's partnership with IBM, a pilot project focused on a specific skillset enabled them to make critical adjustments before deploying their MySkills platform on a larger scale. A proof of concept determines data needs, assesses Al readiness and identifies initial gaps ensuring a smoother expansion.

Analyze Market Data and Validated Skills Frameworks

Identify relevant, in-demand skills using validated databases and frameworks such as O*Net and the Standard Occupational Classification (SOC) System.

Example:

The Department of Labor's O*Net covers over 900 occupations mapped to the Standard Occupational Classification (SOC) system and is built on data from statistically valid samples of workers, ensuring precise skill identification for various roles.

Implement a Dynamic Model for Continuous Updates

Ensure your taxonomy evolves with new technologies, emerging roles, and evolving skill requirements, helping the organization remain adaptable and responsive to future demands.

Example:

IBM and Amazon utilize Al-driven, dynamic skills taxonomies as part of their talent strategies. IBM offers personalized learning paths to stay aligned with emerging technologies, while Amazon fosters continuous learning and skill development to enhance employee agility and mobility

Tailor the Skills Taxonomy to Your Organization's Context

Ensure relevance by customizing the taxonomy to reflect the specific needs and nuances of your organization or industry.

Example:

Research has shown that taxonomies tailored to organizational needs are more sustainable and reflective of ongoing market shifts.

Involve Human Subject Matter Experts

Use human expertise to validate Al-generated taxonomies, checking for accuracy, context, and potential biases in Al outputs.

Example:

Studies show that human intervention is critical to mitigate errors and biases in Al systems. As Al becomes central to HR systems and workforce planning, the need for human Al collaboration remains paramount.

Capture a Comprehensive Range of Skills

Build a taxonomy that includes both technical (core and adjacent) and durable (soft) skills, creating a holistic picture of role requirements and facilitating better talent matching and career development.

Example:

Research shows that organizations that map a broad spectrum of skills gain a competitive edge by adapting quickly to shifting demands, as shown in workforce studies.

Align the Skills Taxonomy with Development Initiatives

Use the taxonomy to guide learning and development programs, mapping identified skills to targeted upskilling and career paths. This alignment clarifies growth opportunities for employees and supports effective reskilling.

Example:

Research highlights that aligning taxonomies with learning initiatives improves skills gap resolution by linking employee development directly to skill demands.

Recredential Job Descriptions for Skills-Based Hiring

Shift from traditional degree requirements to skills-focused job descriptions, increasing inclusivity and better matching candidates to roles.

Example:

OneTen's research indicates that skills-first job descriptions outperform traditional, degree-focused descriptions by improving candidate understanding of qualifications, application likelihood, and perceived fit.

Set Ethical Guidelines and Ensure AI Transparency

Establish clear ethical guidelines to ensure transparent, fair, and trustworthy Al outputs.

Example:
Johnson & Johnson introduced Al ethics guidelines for their MySkills platform and transparency measures to build trust and foster buy-in while allowing employees to opt-out or challenge skill inferences.

This checklist provides a comprehensive, evidence-based guide for implementing an AI-driven skills taxonomy. Ultimately, combining AI technology with human oversight is essential in enabling organizations to develop a future-ready, skills-first workforce that is both adaptable and responsive to evolving market demands. For more detailed findings, read OneTens Research Snapshot: Skills-First Transformation: Building a Skills Taxonomy Leveraging AI.