Role, competency, KRA, KPI — and the one layer you're missing

You built the whole structure — except the layer that makes it usable

If your organisation already works in role → competency → KRA → KPI, you've done more than most ever manage. You've defined the jobs, the capabilities each job needs, the outcomes it owns and the metrics that track them. You've written down what "good" looks like. That is the hard part, and it's done.

And yet, for any single employee, you probably can't answer one basic question with evidence: how good are they, really, at the competencies their role requires — and how do you know? Today that answer is a manager's opinion, written once a year in an appraisal. There is a layer missing between the framework you built and the decisions you make on top of it. This is about that layer.

You defined what good looks like. You never built the thing that measures it.

The broken link in the middle

Look at your framework end to end and the gap is obvious once you see it:

• Your KRAs and KPIs measure the result — did the person hit the number.
• Your competencies describe the capability the role needs.
• Nothing connects them. There is no measured line from "the competency the role requires" to "how much of it this person actually has."

So competency lives on paper — a definition, a dropdown, a rating a manager fills in from memory. And the KPI, the only thing you genuinely measure, is a lagging signal: by the time it tells you a capability was missing, the target has already been missed, the project is already late, the person has already left. You're steering by the wake.

The layer that's missing: skill, measured

The missing layer has a name — the measured skill. It isn't a new framework; it's the measurement underneath the framework you already have. Each competency breaks into a handful of concrete skills. Each skill has a proficiency level, say 1 to 5. Each role says which skills, at which level. Now, for the first time, every person can be measured against exactly what their role requires.

The only new stage is Skill (measured). Everything to its left, you already own. Everything to its right — a real gap, and an action you can defend — only becomes possible because of it.

Measured on evidence, not opinion

This is where most "competency assessments" quietly fail: the measured number is a self- or manager-rating on a form, once a year. That's a claim, not a measurement — and it inherits every bias, halo effect and recency error in the room.

The missing layer only works if the skill is measured on evidence — a graded assessment, a live interview, real work, an independent review — and reported as two numbers, never one: how good (a level) and how sure (a confidence). A skill only counts once it crosses from claim to proof. That's the difference between "the manager thinks she's strong at forecasting" and "she's a Level 4 — here's the evidence, and here's how sure we are."

What the layer unlocks

Slot one measured stage into your existing model and every downstream decision gets sharper:

• Hiring — shortlist on proven skills against the role's required levels, not résumé impressions. Quality of hire up, mis-hires down.
• Mobility & succession — see who already meets the bar for the next role and fill from inside, instead of paying to hire out.
• L&D — aim training at the specific gaps, then re-measure to prove the level actually moved. Finally defend the budget with data.
• Appraisal — back every competency rating with evidence instead of bias, so it holds up in a promotion or an exit.
• Leadership — answer "is our workforce ready for this?" with a number, live, not a hunch.

You don't rebuild anything

Here's the part that makes it easy to start: the measured layer sits on top of your roles and competencies. Nothing to rip out, no framework to relaunch, no two-year programme. You don't boil the ocean — you pick one critical role, map its competencies to measurable skills, measure the people on evidence, and read the gap. If that one picture is useful, you scale it; if it isn't, you've lost a fortnight, not a mandate.

That pilot is a repeatable template — here's how to run a skills gap analysis on one critical role in two weeks. And because performance and capability are different questions, the measured layer is also what stops you mistaking a context-inflated top performer for a genuinely skilled one.

Want to know how much of what you "know" about your people is already proof — and how much is a rating on a form? Take the 2-minute Skills Readiness check, see exactly how we measure a skill on evidence, or talk to us about slotting the missing layer into the framework you already run.