The skills matrix for education teams

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

Why do education teams need a skills matrix?

DfE school workforce statistics show recruitment and retention pressure that capability mapping helps leaders plan around (Department for Education, 2024). In schools, a skills matrix is not a talent spreadsheet: it is how you prove who can teach and cover each subject, and whether every adult on site is current on safeguarding and the other statutory duties that inspections treat as non‑negotiable.

Most headteachers can describe their team in conversation. That picture is partial, shifts with absence, and rarely survives a Monday morning cover crisis. A matrix makes teaching cover and compliance visible before they become an inspection finding or a class without a qualified teacher.

What is an education skills matrix?

An education skills matrix maps staff against two linked views: teaching and cover capability (subjects, phases, interventions) scored on a shared 0–5 scale, and statutory training currency (safeguarding, first aid, SEND awareness, Prevent, fire safety, data protection) tracked as current, due or expired with dates.

Read across a row to see one teacher's profile. Read down a teaching column to see who can deliver or cover that subject at standard. Read the compliance row to see whether the school is audit‑ready on each mandatory item. Used well, the grid answers three operational questions every week: who can teach this class unsupervised, who can cover at short notice, and is anyone on site without current safeguarding training?

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

The floor is the level at which someone may teach or cover a subject or phase without direct supervision. For most teaching competencies that is Level 3 — capable, consistently to standard, signed off by the relevant lead. At or above the floor, the person may be timetabled or allocated cover independently. Below the floor, they work under supervision or cover only with support 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 developing practice and supervised cover; Level 3 is the "safe to timetable alone" threshold for that subject; Levels 4–5 describe leaders who model practice and develop others. Teaching competence still needs refreshing when curriculum or policy changes — date every score when it changes.

Is being below the floor a failure?

No. An early‑career teacher on induction should sit below the floor on most subjects. A specialist asked to cover maths for the first time might be Level 2 until sign‑off. The matrix records that state so cover is deliberate, not assumed from job title.

Flagged below‑floor cells are a safeguard: they tell the cover coordinator who needs a parallel qualified teacher in the room, and they give the developing teacher a visible path to sign‑off. Document supervision requirements beside independence, the same way clinical teams do on wards.

What does a secondary department matrix look like in practice?

Imagine a seven‑person English department scored on six teaching competencies plus a safeguarding currency column. Teaching columns carry a required floor of Level 3. Safeguarding shows current or expired with renewal date, not a 0–5 score.

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

Cells at or above Level 3 are signed off for unsupervised teaching or cover on that competency. A‑level English has only two people at Level 3+: one absence removes A‑level cover entirely. The part‑time teacher's safeguarding due date is an amber compliance flag — not a teaching gap, but an inspection risk if not renewed before expiry.

How should a head of department use the matrix on Monday morning?

Columns first for cover risk, rows first for people. Start by reading down the coverage row on teaching columns. Thin columns are your escalation risks before you accept new timetable swaps. Then read individual rows for development and supervision.

On the example grid, Teacher B is one skill short on A‑level — A‑level classes need either the head, Teacher A, or supervised support until sign‑off completes. Sam is below the floor everywhere on exam‑level teaching, which is appropriate: the matrix should list who supervises each allocated lesson, not treat an ECT as a hidden substitute for an absent colleague.

The HLTA's zeros on GCSE and A‑level are scope markers, not deficits. Including scope explicitly stops false "department gap" counts and protects role boundaries — exactly the distinction inspectors expect you to make.

What outcomes does an education matrix protect?

A school matrix is not paperwork for its own sake. It supports four non‑negotiable outcomes:

• Child safety and compliance — current safeguarding and statutory training visible before anyone enters a classroom.
• Curriculum delivery — subjects staffed by people genuinely capable of teaching them to standard.
• Cover resilience — confidence that absence will not leave classes without competent teaching.
• Structured development — visible support for ECTs and staff stretching into new subjects.

Ofsted and local safeguarding partners increasingly expect evidence, not assurances. A dated matrix row — level, date, assessor — answers "how do you know?" without scrambling for CPD folders the night before a visit. That alone justifies the admin time if the grid stays current.

Trust boards use the same view to prioritise CPD spend: thin subject columns and amber compliance cells become line items in the improvement plan, not vague "workforce development" intentions that never reach the timetable.

How do you run the first calibration session?

Before scores go live, run a 60‑minute calibration with the head of department, a second experienced teacher, and someone who covers frequently. Bring three real scenarios per contested skill — for example what "GCSE English at Level 3" looks like in your scheme of work versus what requires Level 4. Ask: "What observable behaviour equals Level 2 versus Level 3 today?" Write the agreed sentences into the descriptor row.

Calibration is where education matrices earn trust. Without it, senior staff unconsciously score leniently and newer staff score cautiously, which hides the true cover picture. Publish descriptors beside the grid link or on the shared drive, and revisit them when curriculum or exam board requirements change.

How do you evidence a teaching level before sign-off?

A score should rest on evidence, not assertion. In practice, schools combine:

• Lesson observation and learning walks — against written descriptors for the subject and phase.
• Work scrutiny and assessment data — confirming pupils make expected progress when this person teaches the class.
• Subject leader sign‑off — formal record when someone may cover unsupervised.
• CPD and qualification records — supporting evidence behind the score, not a substitute for classroom competence.
• Compliance certificates — dated proof for statutory items, stored behind the currency cell.

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

What mistakes break education matrices?

No defined floor. Without a floor, "can teach" becomes opinion. Set Level 3 explicitly per subject or phase.

Tracking completion, not currency. Statutory training refreshes on cycles; "attended safeguarding in 2021" is not compliance today.

Treating trainees as cover gaps. ECTs should show below‑floor cells with named mentors, not pressure to fill holes.

Confusing scope with gap. HLTAs and TAs must have out‑of‑scope cells on qualifications they must not hold alone.

Scoring on memory. Calibrate after busy terms when everyone "feels" competent.

Mixing performance and capability. Keep appraisal conversations separate from cover ratings.

What should your first 30 days look like?

Week 1: Agree six priority subjects or phases and your statutory list with renewal rules. Week 2: Pilot‑score the permanent team; mark scope for support staff. Week 3: Calibrate disputed cells with lesson examples. Week 4: Link the matrix to cover procedures and book training for thin columns and amber compliance cells.

By day 30 you should answer, without debate, who may take GCSE English cover unsupervised and whether anyone on site lacks current safeguarding training. If you cannot, the matrix is still a draft.

How do supply staff, split roles, and trust-wide cover fit?

Supply teachers often arrive with unknown profiles. A matrix forces an explicit question before allocation: which cells are signed off today, which require a qualified colleague in parallel, and which subjects are out of scope?

Edge case: a strong supply teacher may be Level 3 on KS3 English but Level 2 on your GCSE scheme until they have delivered observed lessons. Without that granularity, cover coordinators allocate from CV seniority rather than evidence — the pattern that produces classes taught by someone not yet signed off.

Multi‑academy trusts should maintain one row per person even when they work across schools, so sign‑off progress travels with them. Currency on safeguarding must be trust‑standard or explicitly noted where local variants apply.

This guide complements Public sector skills matrix guidance on this site for wider workforce context; this page covers day‑to‑day school implementation.

Which site tools help education teams run a matrix?

• Free 5×5 matrix builder (pilot department)
• Upleashed 0–5 methodology
• Descriptor generator for teaching skills
• Skills audit checklist (pre-rating)
• Capability gap ROI calculator
• Minimum standards of capability

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

Define each level in observable classroom behaviours before anyone scores. Anchor Level 3 to signed‑off, unsupervised teaching or cover to your scheme of work standard.

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 sector-ready wording.

Where should you go next on this site?

Keep education.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.

Spreadsheet-first teams can use the Excel Skills Matrix Template (£199) for floors, heat maps, and coverage counts on the same scale. When updates need dates and reminders, PulseAI carries the grid into year one for £1.

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

Frequently asked questions

Should teaching skills and statutory training sit on the same matrix?

Yes, on one grid with clear labels. Teaching columns use the 0–5 scale for cover capability; statutory items track current, due or expired with dates. Splitting them across files is how lapses hide until inspection.

Where should a primary or secondary school start?

Pick six to eight subjects or phases you cover most often, plus your non‑negotiable statutory list. Pilot one department or key stage for a term, calibrate descriptors, then expand.

How do supply teachers and trainees appear on the grid?

Give them their own rows with dated scores. Supply staff may be strong in one subject but not signed off on your local protocols; trainees should show below‑floor cells with named supervisors, not blank gaps.

How often should schools refresh matrix scores?

Quarterly as a minimum for teaching capability. Statutory currency should be checked monthly because certificates expire on fixed cycles independent of your review calendar.

Can we use the Excel template in a school setting?

Yes. The £199 template implements this 0–5 method with heat maps and training outputs. PulseAI automates the same scale when you need dated reminders across a multi‑site trust.

How do we keep ratings fair across teachers at different career stages?

Use the same observable descriptors for everyone on a subject, calibrate with real lesson examples, and keep performance management conversations separate from capability scores.