A proficiency scale a team can rate by

The scale

Rate what someone can be seen doing today, not their title or years in the job. Every cell in a Kunska matrix uses these five levels, and the same words appear everywhere a rating is shown: the matrix, its legend, the development plans and the AI reports.

0 None

Hasn't done it yet.

1 Beginner

Does it with guidance or a checklist.

2 Competent

Does it alone on routine work.

3 Proficient

Handles the unusual cases; others rely on them.

4 Expert

Sets the standard and teaches it to others.

A skill someone wants to learn can sit in the matrix unrated as a growth target. It is as useful to see as a rating.

Where the levels come from

The scale is not invented for Kunska. It combines two well-established models, each chosen for what it does best.

• The names come from the Dreyfus model of skill acquisition (Stuart and Hubert Dreyfus, 1980): Novice, Advanced beginner, Competent, Proficient, Expert. It is the most widely used account of how a person gets better at one skill, moving from following rules to recognising situations to acting on judgement. Kunska folds the first two stages into Beginner, because a team map rarely needs to tell them apart.
• The definitions come from the ILUO skills matrix used in lean manufacturing to plan cross-training: in training, can do it with support, can do it alone to standard, can teach others. Each level describes something you can observe rather than a degree like “intermediate”, which is what makes two people rate the same way.

Frameworks such as SFIA and the European e-Competence Framework (EN 16234-1) are often mentioned alongside these. They describe the responsibility of a role (how autonomous, how much influence) rather than how well someone does one skill, so they suit job architecture and HR reporting better than a cell in a team matrix.

Rating well

• Rate one skill at a time. Expertise is specific: an Expert in one thing is often a Beginner in the next row.
• Use the definition, not the label.Ask “does this person handle the unusual cases, and do others rely on them?” rather than “are they proficient?”
• Expect self-ratings to be off. People new to a skill tend to rate themselves high, and experts tend to rate themselves low. The feedback in a Market of Skills session, where teammates name skills a person left out, is there to correct that.
• Expert means someone teaches it. If nobody on the team could teach a skill, it has no Expert yet, and that is worth knowing.
• Three levels are not enough; ten are too many. Beginner / Intermediate / Advanced leaves the middle so wide that everyone lands in it. Five observable levels are the fewest that separate “can do it alone” from “others rely on them”, and that distinction is the one staffing decisions turn on.

The scale is one part of a skills matrix; the other is who fills it in, and how. That is where most matrices go wrong.