If you and your organization have decided to develop job competency models or a competency-based human resources system, your plan should include the answers to three key questions that will affect the outcome of the project.
- What resources do you have to build your models?
There are many ways to model competencies. Some are complex, time consuming, and expensive. Others are not. Trade-offs are about validation and thoroughness, although less complex approaches may include a validation step.
If you are doing more than one model, you should consider using an integrated approach that uses a common set of core competencies, customizable for each job. Each model requires five to ten days of internal or external consultant time, including facilitation of a high-performance focus group, interviews, and model development.
To get started, choose an outside consultant who is willing to transfer their methodology to you and train your staff to carry out the work, and / or train them to build Competency Models from scratch.
For a large retail organization, a consulting firm made the first two models and certified an internal human resources manager to make additional models. The HR manager also designed and implemented model-based selection and performance review applications. Structured interview questions were developed for each key position to help hiring managers evaluate and select candidates with the required competencies. Performance goals and results forms were also developed.

- Where should you start in the organization?
The best way to demonstrate the benefits of a competencies approach is to start with a high-impact job or one that requires attention, i.e. high turnover, impact on company sales, etc. Define the measurable results of modeling and specify applications.
For example, if you want to model a software developer position, include an application of a recruiting system and an interview guide that will allow you to broaden your pool of candidates and select top performing software developers. Other applications can be added, but you must start with at least one visible and measurable result for the model. If results and applications are not integrated, competency modeling can be perceived as a human resources exercise with no benefits.
There is a natural tendency to want to start with a low-risk, low-visibility position, sometimes to evaluate the process and the consultant. Better to do your homework and thoroughly check references before selecting a consultant than to squander the opportunity to make an impact that can be multiplied throughout the organization.
The ideal place to start is with a director or executive position. Getting that group to develop a model for your position ensures full acceptance. You’ve probably already been through some strategic planning exercises that included thinking about the “core competencies” of your organization. Developing a model validates or alters competencies so that the “ideal” competencies are in fact the competencies required for superior performance in the organization. It also helps them understand the job competition process and align it with the company’s strategy. For example, if innovation is a desired core competency, then a “fostering innovation” competency can be included in most models to drive the exchange rate needed. An executive model is also needed for a good succession planning system.
This is how a large manufacturing division launched its effort to improve performance and alter its culture. A model was developed for division general managers and then cascaded to other key positions.
- Should I make single models for all or multiple models for multiple jobs?
Some organizations use a generic model for all management positions (single model for all). The model may have been developed externally to cover all management positions in all industries. Or it may have been developed in-house by surveying senior executives asking what they believed were the key characteristics required for success in their organization. Both approaches are inexpensive to adopt.
