Insurance Case: Skills Taxonomy Linking Assessment and Training

As roles are reshaped at speed, many organizations face the same challenge: rebuilding the standards they use to look at their people. Company C, a major insurer, started from that same question.

The company focused on its marketing and CRM job family. These roles plan and run the customer journey across policy acquisition, retention, and claims.

The job family included about 100 employees and determined the company’s competitiveness. Yet Company C could not explain, on a consistent standard, which capabilities these roles required or how those capabilities should be developed.

‍

Job Descriptions Without Shared Competency Standards

A closer look at Company C revealed three interlocking problems.

1. No standard for competencies.‍

Job descriptions existed, but they did not capture the capabilities the teams actually used across policy acquisition, retention, and claims. Traditional job analysis was rigorous, but it took several months. By the time the standards were finished, the roles had changed again.
Using an off-the-shelf skills list as-is was no answer either. Capabilities from industries unrelated to insurance came mixed in, which made the list unreliable as a standard for Company C’s own roles.

Assessment results were hard to trust.

With competency standards unsettled, Company C had no consistent, objective way to measure individual competency levels.

Training was disconnected.

Training ran independently of assessment results, so the company could not put numbers on what employees lacked or what had improved afterward.

Hiring, placement, and training investment therefore came down each time to the experience and intuition of a small number of people, and those calls were hard to justify to executives with evidence.

All three problems traced back to one root cause: the standards were not aligned. Competency definition, Competency Assessment, and training were each running on a different foundation.

‍

One Standard for Competency Definition, Assessment, and Training

One root cause meant one solution. Company C placed all three activities, defining competencies, assessing people, and closing gaps through training, on the same data foundation. Here is how it worked, step by step.

1. Building an Insurance-Specific Skills Taxonomy from Global Job Data

The first problem, the standard for competencies, was rebuilt from data.

Telta used LLMs to analyze publicly available job data from leading global insurers including Allianz, MetLife, Prudential, and Zurich, and quickly produced an initial Skills Taxonomy suited to the insurance industry.

Telta then combined that draft with Company C’s own role definitions and terminology. Job descriptions served as the starting point, while interviews with people working in the roles surfaced day-to-day realities the documents did not show.

Rather than importing job data built elsewhere as-is, Telta redefined the skills around Company C’s business context of policy acquisition, retention, and claims.

The result was not a list of abstract competency keywords. It was a Skills Taxonomy Company C could put straight to work in its decisions.

Telta organized more than 100 skills across four teams, 12 roles, and three job levels, from individual contributors to leaders.Each skill was classified as role-specific expertise, a method such as analysis or a framework, or proficiency with a practical tool.

Within the same role, what mattered most shifted by level. In a data analytics role, for example, SQL and database proficiency was the priority for individual contributors, while leaders needed the ability to assess whether an analytics initiative was worth pursuing.

Company C now had a consistent answer to its first question: what does this role require?

2. Measuring Individual Competency Levels Through Simulation-based Assessment

Once the standard was built from data, assessment and training, previously each on its own track, could follow the same standard.

Building on the predefined Skills Taxonomy, Company C selected assessment methods to fit the purpose and the role, with consistency and job relevance as the first priorities. It used Simulation-based Assessment, which presents situations close to the actual work, alongside Behavioral Event Interview (BEI), which verifies how capabilities showed up in past experience.

Telta’s AI scored the written responses against a consistent Scoring Rubric. Individual competency levels that had been invisible until then became visible as scores.

3. Designing the Training Each Person Needs, Based on Assessment Results

The third problem, training, connects naturally to the assessment results. Where the results exposed a skill gap, Telta designed training to close it together with specialized training partners. This was not a reasonable-looking pick from a catalog of courses. It was training aimed at the specific gaps the assessment had revealed.

Through this sequence, Company C completed a single flow from competency definition to assessment to training. With the structure in place, the company could reassess against the same standard after training and see what had changed and by how much.

Set up this way, training no longer ends at “completed.” It can be read against a different question: did the competency actually move? Competency definition, assessment, and training finally connect on one axis.

‍