How to run a skills gap analysis on one critical role in two weeks

Don't boil the ocean. Boil one role.

"We should map our skills" is where most skills programmes go to die. It sounds like a two-year, whole-org, taxonomy-first project — so it gets scoped, deferred, and quietly shelved. The way to actually start is the opposite: pick one critical role, run a real skills gap analysis on it in two weeks, and let the result make the case for everything that comes after.

This is the pilot we recommend to every HR team that hasn't started. It's small enough to finish, real enough to matter, and it produces the one artefact that changes the conversation with your leadership: a gap heatmap — what the role needs, what your people actually have, and exactly where the distance is.

You don't need a skills strategy to begin. You need one role, two weeks, and evidence.

What a skills gap analysis actually is

Strip away the jargon and it's a subtraction:

• The required level — the proficiency the role needs in each skill, on a defined scale (say 1–5).
• The measured level — what each person can actually demonstrate, on evidence.
• The gap — the distance between them, per skill and per person.

The whole exercise lives or dies on that middle number. Measure it with a manager's opinion on a form and you've built an expensive guess. Measure it with evidence — a graded task, an interview, real work, an independent review — and you've built something you can defend in a promotion, a restructure or a board meeting. Everything below is designed to get that second number honestly, fast.

The two-week plan

Week 1 defines the target. Week 2 measures against it. Keep the scope brutal: one role, its core skills, and the handful of people who are in it or feeding into it. Resist every urge to widen.

Week 1 — Define the target

1. Days 1–2 · Pick the role. Choose on criticality, not convenience: revenue impact, operational risk, hard-to-hire, or a succession bench that's dangerously thin. If it broke tomorrow, would the business feel it? That's your role.
2. Days 2–3 · List the skills it actually needs. Eight to fifteen, drawn from the work — not fifty scraped from a generic library. Sit with a strong incumbent and their manager for an hour; the real list is shorter and sharper than any framework.
3. Days 3–4 · Set the required level per skill. For each one, define what "good enough for this role" means on your 1–5 scale. This is the target line every person will be measured against.
4. Days 4–5 · Decide how each skill is evidenced. Assign a measurement method to each skill — a graded assessment, a structured interview, a work sample, a 360 review. The rule: nothing counts as "measured" if the only evidence is someone's say-so.

Week 2 — Measure and read

1. Days 6–8 · Measure the people. Run the assessments and interviews, pull the work evidence, collect the independent input you defined in Week 1. You are gathering proof, not asking people to rate themselves.
2. Days 8–9 · Score to level and confidence. Two numbers, never one: how good (the level the evidence supports) and how sure (how much independent evidence stands behind it). A single item is a lead, not a verdict.
3. Days 9–10 · Build the gap heatmap. Required vs measured, per skill, per person — plus a role-level rollup. The distances jump off the page.
4. Day 10 · Read it and decide one thing. Don't try to fix everything. Make a single, better decision on the strength of the map — that's how a pilot earns its sequel.

What the heatmap looks like

Here's an illustrative read for a Data Analyst role — the required line against where the team actually lands:

Two things become obvious in ten seconds: the role is exposed on data modelling and experiment design, and it's over-invested in dashboarding. That single picture reframes hiring, training and who's ready to step up — none of which the org chart could tell you last week.

Read the heatmap: three moves it unlocks

• Hire or deploy to the real gap. Stop writing job specs from last year's template. The gaps are the spec — and if someone elsewhere in the org already measures high on them, you may not need to hire at all.
• Develop with a target. Point training at the two skills that are actually short, not the catalogue — then re-measure to prove the level moved.
• See who's ready to move. A person who meets or exceeds the required line is a mobility or succession candidate you can defend with evidence, not a hunch.

Five traps that sink a pilot

1. Self-ratings. Ask people to score themselves and you've measured confidence and optimism, not skill. Use evidence.
2. The generic library. Fifty off-the-shelf competencies give you a tidy list and a meaningless gap. Fewer, role-true skills beat a comprehensive fiction.
3. Measuring everything. Coverage is the enemy of finishing. One role, its core skills, done in two weeks, beats a half-mapped org that never ships.
4. The big-bang launch. Announcing a company-wide skills initiative before you've proven one role is how programmes collapse under their own expectations.
5. Deciding on level alone. A level without its confidence is half a number. Keep the two together on every decision.

From one role to the org

The pilot is deliberately small, but it's a template. Once one role is mapped and one decision is better for it, the next role is faster — same scale, same evidence bar, same heatmap — and the roles start sharing skills, so your measurements compound. Ten roles in, you're not running pilots anymore; you have a live skills map. That's how a skills-based organisation actually gets built: not with a two-year programme, but with a second role, then a third.

Not sure this is the right first move for where you are? The 2-minute Skills Readiness check places you on the claim-to-proof curve and tells you whether a one-role pilot is your next step. For how the measurement itself works — level, confidence and the claim-to-proof line — see how we measure, or talk to us about running the pilot on a role that matters.