Behavioral Competency

Technical competencies describe what someone can do: operate a system, close a ledger, write a contract, run a payroll cycle. They are demonstrated by doing the thing, and they are relatively easy to assess because the output is visible.

Behavioural competencies describe how someone works: how they respond when a plan fails, whether they raise a problem early or late, how they behave when they disagree with a decision, how they treat people with less authority than them.

The distinction matters because the two predict different things. Technical competence predicts whether someone can do the current job. Behavioural competence predicts rather more about whether they will do it well in a team, and considerably more about how they will fare in a different or larger role.

It also matters because they are assessed differently. Technical competence can be tested. Behavioural competence has to be inferred from evidence of past conduct, which is why the definitions have to be concrete enough for evidence to attach to them.

The single most common defect in competency frameworks is defining a competency as a personal quality.

The left column cannot be assessed, only felt, which means assessments reflect the assessor. It is also where bias enters most easily, because a judgement about whether someone has presence is largely a judgement about whether they resemble people the assessor already regards as senior.

The right column can be evidenced, discussed, disagreed with and improved. That is the whole test: could two assessors independently point at the same evidence and reach the same conclusion.

Frameworks usually define levels, and usually define them badly, by repeating the same behaviour with escalating adverbs. Consistently does X, then always does X, then role-models X.

Useful levels describe different behaviour, not more of the same. A progression for raising risk might run from raising risks in their own work, to identifying risks in the team's work before others do, to changing how the team surfaces risk so that it happens without them.

The test is whether someone could be clearly at one level and not the next, on evidence. If the levels differ only in adverbs, assessors will place people by overall impression and then select the wording that matches, which is assessment in reverse.

Most organisations have more competencies than anyone can hold in mind, usually because the framework was assembled by collecting suggestions rather than by deciding.

A framework with twenty competencies is not assessed. Assessors pick the three they understand, or rate everything at the midpoint, and the resulting data has the appearance of precision and none of the content.

• Six to eight organisation-wide competencies is a workable ceiling for most organisations.
• Add role-specific behaviours where genuinely needed, rather than expanding the common set.
• Cut anything nobody can produce an example of. If no assessor has ever cited evidence against a competency, it is decorative.
• Check that the set actually distinguishes people. A competency where everyone scores the same is measuring nothing.

The last check is worth running on existing frameworks. It usually eliminates a third of them in one pass.

Behavioural definitions are written by people and carry their assumptions about what good conduct looks like.

Wording built around assertiveness, visible confidence, challenging openly and speaking up in large groups describes one register of professional behaviour. It is not a universal one, and it maps unevenly onto people from different cultural backgrounds, people with less positional authority, and anyone for whom directness towards seniority carries a different cost.

The effect is that a competency intended to measure contribution partly measures conformity to a communication style, and the people it disadvantages are predictable.

The correction is to define the outcome rather than the register. Ensures their view is considered before a decision is taken is achievable by several routes, of which speaking up in a large meeting is only one. Speaks up in meetings is achievable by one route.

This is worth checking directly rather than reasoning about: look at competency scores by group. Where one competency shows a pattern the others do not, the wording is doing something other than what it says.