A reviewable tutoring goal has six parts: a baseline, one observable capability, tutor and learner actions, comparable evidence, a review date and clear limits. “Improve maths” or “be more confident” is not enough to plan or judge tuition.

Complete the goal agreement

1. Baseline

Use a real task connected to the current need. Record the work produced, accuracy/quality, explanation and help used.

“On two ratio questions, the learner sets up the relationship correctly but needs a prompt to scale both quantities and cannot explain why the operation applies.”

A grade alone can combine many topics and conditions.

2. Observable capability

Describe what should become more accurate, clear or independent:

“Given a fresh ratio problem, the learner will select and explain an appropriate scaling method and complete it without step-by-step prompts.”

Avoid guaranteeing a final grade. A grade may be a later outcome to monitor, not a precise small target.

3. Shared actions

PersonAction
Learnerattempt, explain thinking, complete agreed bounded practice, raise questions
Tutordiagnose from work, teach, fade prompts, provide fresh checks, record evidence
Parent/carersupport attendance/setup, maintain boundaries, review without doing work
School, if involvedclarify course/feedback/support within its role

4. Comparable evidence

Choose fresh work requiring the same skill, not the memorised original. Record help used. For essay work, compare the same criterion on a different question. For maths, use differently presented questions requiring the same method.

5. Review date

There is no universal number of sessions. Choose a date that allows a reasonable sample for this goal and budget, then decide earlier if safety, conduct or a clear mismatch requires it.

6. Limits and referrals

Record what tuition does not cover and what triggers a teacher, school support team or qualified specialist conversation. An ordinary tutor should not diagnose learning differences or replace safeguarding/health support.

Use the one-page template

Specific need and course:
Baseline task/evidence:
Help currently used:
Observable capability:
Learner action:
Tutor action:
Between-session practice:
Fresh review evidence:
Review date:
Limits/referral triggers:
Decision options: continue / adjust / connect / change / stop.

The DfE tutoring review examined school programmes, not private family contracts. It found stronger assessment systems used formative evidence linked to what pupils were learning and could help determine when tuition was no longer needed. Treat this as a useful design principle, not a private-tuition guarantee.

Check the learner understands the goal

Ask the learner to explain:

  • what they are working on;
  • why it matters;
  • what they will try;
  • how they can ask for a different explanation;
  • what evidence will be reviewed.

If the goal belongs only to adults, engagement and interpretation become harder.

Use the first tutoring session checklist to collect baseline evidence and the progress review at the agreed date. The private tuition guide places the agreement in the full decision cycle.

Build the goal with all three voices

Ask separately:

  • Learner: what task feels difficult, what help is acceptable and what would feel meaningfully different?
  • Parent/carer: what pattern is visible, what school information exists and what practical limits matter?
  • Tutor: what the current evidence supports, what is teachable and what cannot yet be concluded?

Differences are useful. A parent may focus on grades, a learner on panic when starting questions and a tutor on one missing prerequisite. The agreement should connect them without pretending they are identical.

Write the five-part agreement

PartExample
BaselineOn three fresh linear-equation questions, learner completes the first step but needs a specific prompt when the unknown appears on both sides
CapabilitySelect and carry out the rearrangement independently
Teaching/practiceTutor models one comparison, learner explains decisions, prompts are faded on fresh questions
EvidenceComparable questions, support level, later mixed check and relevant school work
ReviewFamily and tutor review after the agreed initial period; continue, adjust, connect or stop

Avoid grade guarantees. A grade can be a later outcome measure, but it combines content, assessment conditions and many skills.

Test whether the goal is teachable

The goal is ready when:

  • it names a bounded course-relevant task;
  • the baseline comes from observable work;
  • the tutor can explain the next teaching decision;
  • the learner can attempt a comparable task;
  • support can be recorded and reduced;
  • the family knows what happens at review.

It is not ready if the main issue is still unknown, a school requirement is being guessed, or a safety/health/specialist concern requires another route.

Use leading and later indicators

Early evidenceLater evidence
Learner can explain the task and goalSkill appears in current school work
Tutor identifies a precise prerequisiteFresh tasks are completed with less support
Practice is aligned to courseLearning survives a gap or mixed context
Feedback creates a next actionAssessment evidence changes where comparable
Learner can ask a precise questionDependence on tutoring decreases

Do not expect all later outcomes after one or two sessions.

Copyable tutoring-goal agreement

Learner/course/board/tier:
Learner's view:
Parent/school evidence:
Baseline task and support level:
Observable capability:
Tutor teaching approach:
Learner practice responsibility:
Evidence to collect:
Communication route:
Review date:
Continue/change/stop conditions:

Sources

  1. Department for Education — Independent review of tutoring, phase 1 (2026-07-23)
  2. Pearson — Planning and delivering tutoring in maths (2026-07-23)

Keep reading