Steven Forth is a Co-Founder of TeamFit. See his Skill Profile.


Competency models are a powerful way to organize jobs, roles and skills. They can be used as a lens to get insights into the skills in an organization (or a team or an individual) and to answer key business questions.

  • Do we have the skills we need to win?
  • What potential skills could we develop?
  • How are we applying the skills we have?
  • Where do we have gaps? Today and tomorrow.
  • Where do we need to invest in development?

A good competency model is unique to an organization and its goals. It is a way to develop strategic differentiation. Given this, what goes into a good competency model and how does one go about building one? Here is a simple step-by-step process.

  1. Set goals
  2. Define uses
  3. Gather resources
  4. Analyze resources
    • Bottom up approach
    • Top down approach
  5. Create the framework
  6. Populate the framework
  7. Connect to people and work

Before diving in, let’s take a high level view of how the pieces fit together. The competency model sits at the centre. It is where all the different inputs come together and get connected. This can be complex process and a lot of work, so we have built an AI to support this. It lets you upload existing documents, parses them to find jobs, roles and skills, and helps to populate the competency model.

Once built, the competency model provides a compelling lens to look into the organization. When combined with skill profiles (actual data on individuals, projects and teams) the competency model answers the five key business questions. This is not the end of the story though. Given the critical role that skills play in success it is important to have a system of record for skills (see Do you need a system of record for skills?). At this point in time, other enterprise applications have only rudimentary skill records. These records are usually no more than a simple list of skills, sometimes in some form of hierarchy. This is not robust enough to share data across the ecosystem. Skills data needs to include information about the type of skill, level of expertise and how it connects to other key business objects. Some of the applications that need skill data to function effectively are the applicant tracking systems (ATS) used in the hiring process, the various systems used in learning and development (the learning management system or LMS, the learning experience platform or LXP, the learning record store or LRS). Performance management systems (PMS) and the central talent management or human resources information system (HRIS) are also consumers of skill records.

Getting back to the process, here are some simple steps to follow.

1. Set goals

Competency models are powerful tools and can serve many purposes, but trying to achieve too many goals at once can lead to confusion. Choose just a few goals for your initial implementation and make sure they are aligned with strategic imperatives. Example goals could include …

  • Aligning skills with current needs
  • Aligning skills with future needs
  • Increasing employee engagement
  • Increasing staff utilization
  • Rationalizing learning resources to fit skill development needs
  • Understanding which skills are driving business outcomes

2. Define uses (connect goals to outcomes)

Ask how the competency model will be used and who will use it. These uses should of course map back to the goals that have been established. As part of defining use, it helps to identify the users. Begin by asking who will be using the competency model and how will they use it. Is it a model for executives, for operational leaders, for HR, for learning and development, or is it primarily to be used by individuals in understanding and building their own skills? Will the model be used directly or will it be used to organize other systems? Use cases can help with this, but you will want to make sure that you connect the use cases to goals and to outcomes.

3. Gather resources

Many different resources can feed a competency model. The obvious ones are things like organizational design documents, job descriptions, career paths and so on. There can also be a lot of useful information in the learning development organization. Learning plans, instructional design materials, and course outlines can all have relevant information on skills.

It is important to go beyond this though. Look at actual work products to more fully inform an understanding of the skills being used in specific jobs and roles and how they are impacting business outcomes. Consider including white papers, patents, project reports and data from project management systems (there is a clear mapping from tasks to skills).

Performance reviews can also have useful information about the goals and key performance indicators for jobs and where people are excelling or struggling. This can give powerful insights that can inform the competency model.

4. Analyze resources

As you gather these resources, you have to analyze them. The goal here is to understand what skills are being used in each job or role and how these are contributing to success.

If you have access to a skill profile and management platform, like TeamFit, you can accelerate this work by using information about people’s actual skills and how they are using them. TeamFit organizes skills into general categories (Foundational, Business, Social, Technical, Tool, Design, Domain) that can help inform design, and you can also see collections of skills by role or project type.

There is a risk of being too focussed on the present state. A good competency model should also be a guide to the future. Make sure that you consider future skill requirements and emerging job roles.

Given the amount of data to be integrated in this process, some level of automation is a big help. TeamFit provides this through its AI which can extract skills and other constructs from unstructured data. This makes it possible to effectively consider a much wider range of content and can provide important insights.

5. Create the framework

There are many ways to organize competency models. We are developing a library of competency model design patterns Which pattern works depends a lot on the goals for the model and how it will be used. Based on our analysis of existing models and surveys, the most common components are listed below. (If you are interested in what came up under ‘Other’ please contact us.)

The design job is to choose and organize the components of the model and define their relationships. Will your model include Jobs or will it begin with Roles? Are Jobs composed of Roles? Are there certain skills that apply to all Jobs or Roles for your organization? How about behaviours? Many organizations include the behaviours associated with their core values in their competency models.

Another point to be considered is a skill category system. TeamFit provides the default categories noted above (Foundational, Business, Social and so on) but many organizations need to apply their own categories based on their own models of work, leadership and learning and development.

6. Populate the framework

Once you have built your framework, you will need to build the actual competency model. At TeamFit we take see three basic modes: top down, bottom up, document driven. These are not mutually exclusive. In fact the best competency modellers move back and forth between them whenever possible.

Top down is the conventional approach. Smart people get together (competency model design is almost always collaborative work) to identify the jobs, decompose the jobs into component roles, competencies or behaviours, identify the skills required along with the level of expertise, and build everything back up into a comprehensive model. The model is then reviewed by various domain experts and by the people who will be using it. In our experience, a review by the learning and development team also adds a lot of value, even when the primary purpose is not learning.

Bottom up is the emerging approach in which the skills identified by people during actual work are identified and the model is built from the bottom up by mapping these skills to projects, job roles and ideally to business outcomes. TeamFit Skill Profiles are one way to gather and begin to organize this information. We find it valuable to complement this with skill interviews to get richer insights.

Document driven modelling takes advantage of recent advances in natural language processing and connects them to skill graphs (skill graphs are databases that map the connections between skills and other skills and between skills and people, jobs, roles, behaviours and so on) to enable semantic reasoning. This is how part of the TeamFit AI works. Upload a document and TeamFit parses it for skills and other relevant content that can be used to build a skill model. As with any machine based system, the results require careful human curation.

7. Connect to people and work

Competency models become real when people use them to work and learn. The first step is to help people assess their own skills using the competency model as a lens. Connected to people’s real skills and experience, competency models help to identify skill gaps. Once the gaps are visible, people can move to address them.

Another use of competency models is to assess the coverage of learning resources. Given a competency model, are learning opportunities available to help people improve their expertise. Learning opportunities do not need to be limited to formal learning. Opportunities to work on new projects in new roles, mentoring support, communities of practice, are all ways to support learning.

An emerging best practice is to connect competencies to business outcomes. The basic approach is to …

  1. Design and validate a competency model
  2. See what competencies are shown in specific work products or projects
  3. Record the business outcomes
  4. Map the circuits between competencies and business outcomes
  5. Evolve the competency model to better align with superior outcomes

Competency model design is a rich topic, and as TeamFit roles out its competency modelling environment over the course of this year we will be writing much more on this theme.

How is your organization using competency models? You can share your ideas and experience in this short survey.