The skills matrix for public sector and government teams

Content aligned to the Capability Guide PDF for this topic. Q2 2026 refresh.

Why do public sector and government teams need a skills matrix?

ONS public sector employment stood at 6.19 million in the latest release, a scale that makes role-by-role capability planning essential (Office for National Statistics, 2025). In government and wider public bodies, a skills matrix is not a talent inventory for its own sake: it is how you show whether statutory and operational duties can be delivered with the capability you actually hold—not the headcount chart alone.

Most permanent secretaries and directors can describe their organisation in conversation. That picture fragments across directorates, hides single points of failure on essential services, and rarely survives restructure or spending review. A matrix maps capability against duty so thin cover, consultant dependence, and cross-cutting shortages surface before a service falters or an audit asks for evidence you do not have.

What is a public sector skills matrix?

A public sector skills matrix maps people—or roles and professions—against the competencies each service and directorate needs to discharge its duties. Each cell holds a level on a shared 0–5 scale. Each competency carries a required floor, usually Level 3: capable, consistent, unsupervised delivery to your published standard.

Read across a row to see one individual’s profile across policy, digital, commercial, or front-line skills. Read down a column to see coverage for a competency shared across services. Read against the floor to see where a statutory function rests on one or two people, where development is underway, and where external spend is masking a gap you should build permanently.

Used well, the grid answers three questions leaders ask every quarter: which duties are at risk if key people leave, which professions need shared development rather than siloed courses, and where consultant invoices prove a capability you have not invested in-house.

What is the required floor, and why is Level 3 the usual line?

The floor is the minimum level for unsupervised delivery on that competency—casework, inspection, commissioning analysis, digital service design, or front-line service. At or above Level 3, work may be allocated without a named supervisor for that skill. Below Level 3, tasks stay in supervised development until evidence supports sign-off.

The Upleashed 0–5 framework used on this site makes that line explicit in every cell. Levels 1–2 describe learning and supervised practice; Level 3 is the operational “safe to assign alone” threshold; Levels 4–5 describe experts who handle complexity and develop others. In regulated services, Level 3 must be defined in observable behaviours, not vague “competent” labels.

Competence is not permanent. Policy change, new casework systems, or restructure alter what Level 3 means. Date every score when it changes, and reconfirm skills that have not been practised within your local policy window—especially safeguarding, licensing, and financial delegation.

Is being below the floor a failure?

No. Treating below-floor cells as failure is how matrices become decorative. A policy officer on first casework should sit at Level 1 or 2 on complex assessments until sign-off. A digital apprentice should show developing scores on live systems under supervision. The matrix records that state so allocation is deliberate.

Flagged cells tell service managers who must supervise whom, and give learners a visible path to sign-off. Document supervision in the same place you document independence, so audit trails and development conversations align.

What does a directorate service matrix look like in practice?

Imagine a children’s services directorate scored on six cross-cutting competencies: statutory assessment, digital casework, commissioning analysis, partnership management, court reporting, and quality assurance. Every competency carries a required floor of Level 3 for unsupervised delivery.

*0 = out of role scope (not a gap).

Court reporting and quality assurance each have only two people signed off at Level 3+. One long-term absence removes a statutory function from safe cover. Commissioning has the same pattern—often filled by consultants because the matrix was never used to justify permanent build. That is the conversation the grid forces: build, share, or buy—and on evidence.

How should a service manager use the matrix on Monday morning?

Columns first for duty risk, rows second for people. Start with the coverage row. Thin columns are escalation priorities for training and recruitment. Then read rows for supervision and development.

On the example grid, statutory assessment has three people at Level 3+, but court reporting has only three with two below floor on complex work—plan supervision for high-risk hearings until sign-off completes. Sam is below floor everywhere, which is appropriate: list named supervisors per allocated task rather than treating Sam as a coverage gap. The agency row shows partial competence; do not assign commissioning or quality tasks outside scope.

Where consultant spend is high, compare invoice lines to thin columns. If commissioning analysis is red on the matrix and heavy on external spend, the business case for in-house development writes itself.

What outcomes does the matrix protect for citizens and scrutineers?

A public sector matrix supports four outcomes that translate directly into duty, cost, and trust:

• Statutory delivery — essential services stay resilient before key people leave.
• Public purse — capability gaps visible before they become permanent consultant reliance.
• Cross-cutting capability — digital, data, commercial, and project skills planned across professions.
• Accountability — dated evidence for gap analysis and manager stewardship of development.

Inspectors, internal audit, and elected members increasingly ask “how do you know?” A dated matrix row—level, date, assessor—answers without a scramble before committee or regulator review. That alone justifies disciplined upkeep if the grid stays current.

How do you run the first calibration session?

Before scores go live, run a 60-minute calibration with the service manager, a profession lead (social work, finance, digital), and one team representative. Bring three real scenarios per contested skill—for example what observable behaviour separates Level 2 from Level 3 on statutory assessment after a policy update.

Write agreed sentences into descriptors shared on the team site. Without calibration, senior staff score leniently and newer staff score cautiously, which hides the true cover picture executives need. Revisit descriptors when legislation or system changes alter the floor.

How do you evidence a level before sign-off?

Scores should rest on evidence, not assertion. Public bodies typically combine:

• Supervised casework or delivery — observed practice with a named assessor.
• Professional standards and portfolios — where registrants apply.
• System proficiency checks — for digital tools that gate casework.
• Peer review or audit sampling — confirming quality holds at scale.
• Training linked to practice — attendance plus demonstrated use on live work.

The matrix stores the current level and date; the evidence file sits behind it. That pairing is what survives scrutiny.

What mistakes break public sector matrices?

No defined floor. Without Level 3 (or your local standard) per competency, “competent” becomes opinion and the grid cannot drive allocation.

Mapping job titles only. Duties cross professions; map skills that gate delivery, not grade labels alone.

Treating supervised development as failure. Show below-floor cells with named supervisors for trainees and movers.

Ignoring consultant signals. Persistent external spend on a thin column is a workforce plan input, not a separate budget story.

Building once. Re-score when policy, systems, or structure change; a static matrix becomes fiction within a quarter.

Mixing capability and performance. Keep appraisal separate from competence ratings.

What should your first 30 days look like?

Week 1: Agree six duty-critical competencies and draft descriptors. Week 2: Pilot-score one service; mark scope for roles that do not perform certain tasks. Week 3: Calibrate disputed cells with real cases. Week 4: Link the matrix to gap analysis, training plans, and recruitment for thin columns.

By day 30 you should answer, without debate, which statutory tasks may be allocated unsupervised and what happens to safeguarding cover if your senior social worker is on leave. If you cannot, the matrix is still a draft.

How do agency staff, secondees, and shared services fit?

Agency and secondee rows must show which cells are signed off today, which require supervision, and which tasks are out of scope—not a blanket “experienced” label. Edge case: a secondee strong on digital casework may be below floor on court reporting for your authority; score columns independently before allocation.

Shared services (HR, IT, legal) should appear as columns with coverage counts per customer directorate, so internal SLAs reflect capability, not ticket volume alone.

Which professions and competencies should a public body map first?

Start with competencies that gate statutory delivery, not every soft skill in the corporate framework. Typical cross-cutting columns include digital casework, data analysis for decisions, commissioning and contract management, project and programme delivery, policy drafting with legal clearance, safeguarding escalation, and customer communication for regulated services. Front-line professions add their own: social work assessment, environmental health inspection, revenues and benefits processing, housing allocation, or highways response.

Limit the first release to six columns per pilot service. Each column needs a named owner in the profession who will maintain descriptors when policy changes. If you cannot name that owner, the column is not ready to score.

Layered reporting matters for executives: a service-level view with coverage counts, then team-level grids with named individuals for managers. The executive view answers “is this duty safe?”; the team view answers “who supervises whom this week?”

When two directorates share a scarce skill—commercial analysis or cyber security—maintain one profession column across both rather than duplicating incompatible definitions. Shared columns make internal secondment and cover pools possible; duplicate columns hide the true shortage.

Which site tools help public sector teams run a matrix?

• Upleashed 0–5 methodology
• Descriptor generator
• Skills audit checklist (pre-rating)
• Capability gap ROI calculator
• Minimum standards of capability
• Workforce capacity planning

How should you score skills on the 0–5 scale?

Define each level in observable behaviours before anyone scores. Weighting and full definitions live on the 0–5 scale guide; industry matrices use this summary table.

Capability percentages use Upleashed weightings (Level 1 = 25%, Level 2 = 50%, Level 3 = 75%, Levels 4–5 = 100%; Level 0 excluded). See competency scale 0–5 explained for the full framework.

See the methodology pillar and descriptor generator for role-ready wording.

Where should you go next on this site?

Keep public-sector.pdf for offline briefings. Online, you get searchable structure, tables, and pointers into the wider methodology.

If descriptors drift between managers, reset them against the methodology pillar and republish from the descriptor generator.

A blank sheet works for week one; the Excel Skills Matrix Template (£199) removes formula risk when you add floors and analytics. Later, PulseAI keeps evidence current without rebuilding the model.

Publish descriptors beside the grid so new managers inherit the same meaning of each level, not their own interpretation.

Frequently asked questions

Should a council matrix list every employee or every role?

Start with roles or pooled professions for planning views, then attach named individuals where managers hold development accountability. Large bodies often maintain a service-level heat map for executives and team grids for line managers.

How do we stop the matrix becoming another HR spreadsheet?

Tie each review to decisions: training booked, recruitment briefed, consultant scope reduced, or service continuity plan updated. If the grid does not change a decision within a quarter, you mapped too much or score too rarely.

What is the first competency set to pilot?

Pick one statutory service with known pressure plus three cross-cutting skills that gate delivery—often digital casework, commissioning analysis, or safeguarding escalation. Six columns and one directorate is enough for month one.

How often should public sector scores refresh?

Quarterly minimum; monthly when legislation, systems, or restructures change duties. Reconfirm any skill unused beyond your local policy window, especially for regulated casework.

Can we use the Excel template in government?

Yes. The £199 template implements this 0–5 method with heat maps, required levels, and training outputs. PulseAI automates the same scale when you need dated evidence across departments.

How does this support accountability to members or regulators?

It provides a dated baseline for gap analysis: which duties are covered, which rest on one or two people, and what development closes the gap. That is evidence scrutineers expect—not assertion in a slide deck.