School–tutor partnerships should show what changed

Tutoring happened. The learner attended. The sessions were delivered. Topics were covered. Activities were completed. The tutor provided feedback.

Everyone did what they were supposed to do.

But one question still remains: What changed?

That is the question a responsible tutoring partnership should eventually be able to answer. Not because every educational outcome can be reduced to a number. Not because tutoring should produce instant grade increases. And not because tutors should spend more time producing reports than teaching. But because additional support should exist for a reason.

There was a starting point. We hoped something would improve. The tutoring was intended to help. So at some point we need evidence about whether it did.

Delivery is not impact

This distinction sounds obvious, but education systems often measure what is easiest to count.

Sessions delivered.
Hours completed.
Attendance.
Activities.
Topics.
Questions answered.

All of those can tell us whether tutoring took place. They cannot, by themselves, tell us whether it worked. Twenty hours of tuition is twenty hours of provision. It is not automatically twenty hours of progress.

Likewise, “Covered algebra” does not tell us whether the learner now understands algebra better.

“Completed ten sessions” does not tell us what they can do independently now.

“Engaged well” does not tell us whether the original learning problem changed.

This is why tutoring needs to begin with something more useful than a timetable. It needs a purpose.

^{Tutoring moves from a starting point through targeted support to evidence and the next decision.}

Start by defining what tutoring is supposed to change

“Needs help with maths” is understandable, but hard to evaluate.

Compare it with: “Can apply the taught method when prompted, but cannot yet identify independently which method a problem requires.”

Now there is something to work with.

Or: “Understands the text verbally but struggles to structure that understanding into an extended written answer.”

Or: “Has the required science knowledge but loses marks because they misinterpret multi-step questions.”

These descriptions are not perfect diagnoses. They are working hypotheses, but they give the tutor somewhere meaningful to begin.

And importantly, they make it possible later to ask: Is this still true?

That is the beginning of evidence.

Tutoring should connect to the learning need, not merely the subject

A learner may be receiving maths tutoring. That does not necessarily mean the problem is simply “maths”.

Perhaps the barrier is:

a missing prerequisite;
difficulty interpreting questions;
poor recall;
low confidence when working independently;
an insecure method;
or repeated calculation errors.

The subject tells us the broad area; the learning need tells us what the tutoring should actually address. This is one reason alignment with school matters.

The Education Endowment Foundation's evidence on one-to-one tuition says tutoring is more likely to have an impact when it is additional to and explicitly linked with normal teaching.

Its implementation guidance also recommends understanding learning gaps, connecting tuition to classroom content, creating opportunities for teacher and tutor communication, and monitoring progress so provision can be adjusted.

That is more than curriculum coordination.

It creates a loop:

What does the learner need?
What is the tutor doing about it?
What evidence is emerging?
What should happen next?

A partnership needs a baseline

“Baseline” can sound like a formal assessment. It does not always need to be.

A useful starting point might be:

a recent piece of work;
a small diagnostic task;
teacher feedback;
an examination question;
an example of homework;
or an observation from the first tutoring session.

The purpose is not to generate another score.

It is to create enough evidence to say: This is what the learner can currently do.

And: This appears to be where the difficulty begins.

Without that starting point, it becomes surprisingly hard to know whether something later improved. A learner may feel better. The parent may be pleased. The tutor may report positive sessions. All useful, but compared with what?

The evidence should match the objective

Suppose tutoring is intended to improve independent problem-solving. A higher score achieved with extensive tutor prompting may not be the evidence we need.

We might instead look for:

fewer prompts;
better choice of method;
greater willingness to attempt unfamiliar problems;
or successful application when the tutor is not present.

If the goal is essay structure, useful evidence might be whether the learner can now:

identify the task;
plan an argument;
select evidence;
and organise paragraphs independently.

If the aim is retrieval of key knowledge, we may look at what can be recalled after a delay.

If the objective is confidence, we should still make that observable.

Not simply: “Confidence has improved.”

But perhaps: “The learner now attempts unfamiliar questions before asking for reassurance.”

The evidence needs to tell us something about the change we were actually trying to create.

Progress does not always appear first in grades

This matters. Imagine a learner predicted a Grade 4 begins tutoring in September. Three weeks later, their predicted grade is still 4.

Has tutoring failed? We cannot know from that information alone.

Perhaps they now understand the foundational topic that had been blocking later work.
Perhaps they can begin questions independently.
Perhaps they are making fewer repeated errors.
Perhaps they understand feedback better.
Perhaps none of those things has changed.

The grade is useful, but it may be too slow or too broad to explain what tutoring is doing right now. Grades are often lagging indicators.

A good partnership looks for smaller pieces of evidence that show whether the learner appears to be moving toward the larger outcome. That does not mean replacing attainment with softer measures. It means understanding the path between the two.

The strongest evidence may be surprisingly small

We do not necessarily need a complicated dashboard.

Imagine the original problem was:

Learner can solve simultaneous equations when told which method to use but cannot choose a method independently.

Four weeks later, the tutor records:

Learner correctly selected an appropriate method in four unfamiliar examples without prompting. One error resulted from calculation rather than method selection.

That is useful evidence; something changed. We also know something remains.

The next decision might be: reduce support further and introduce more varied problems.

Or perhaps the tutor discovers:

Learner can now choose a method during tutoring but did not transfer this successfully to recent schoolwork.

That is also valuable. It tells us the tutoring has produced some progress, but the learning has not yet transferred reliably. The next decision changes. That is the purpose of evidence.

^{Tutor and learner comparing earlier and recent work to identify a specific improvement.}

Evidence should allow us to change course

Monitoring progress has little value if the programme continues unchanged, regardless of what we discover.

The learner improves quickly. Change the challenge.
The learner remains stuck. Change the explanation.
A missing prerequisite appears. Go backwards.

The original problem turns out not to be the real problem. Redefine it.

The learner can perform successfully with the tutor but not independently.
Reduce prompting and test transfer.
Nothing appears to be improving.
Ask whether the current tutoring approach is appropriate.

This is one of the strongest reasons to make progress visible. It creates permission to adapt.

EEF's tutoring guidance explicitly recommends monitoring pupil progress and adjusting provision accordingly.

Monitoring is not an administrative endpoint; it should change something.

Teacher and tutor do not need to share everything

A responsible partnership does not require constant communication. Teachers already have enormous workloads, tutors do not need unrestricted access to school data, and schools should not automatically share personal information with external tutors simply because it might be convenient.

The goal is not maximum information. It is minimum useful information.

At the beginning, the tutor may need:

the current curriculum area;
the specific learning concern;
relevant recent evidence;
and the intended outcome.

During tutoring, the school may only need to hear something if it would change classroom action. For example:

Useful to share

Learner understands the concept when represented visually but continues to misinterpret text-heavy versions of the same problem.

Less useful:

Completed Lesson 6 and worked hard.

The filter should be: Would this information change somebody's next decision?

If yes, it may be worth communicating.
If not, it may not need to travel.

Parents should not be the only evidence bridge

In many tutoring arrangements, the parent becomes responsible for connecting the system.

The school says something.
The learner remembers part of it.
The parent explains it to the tutor.
The tutor sends the parent a summary.
The parent decides whether anything should return to school.

Families should remain central, but they should not have to manually reconstruct the educational evidence on behalf of professionals.

Where an appropriate school–tutor relationship exists, even a very small agreed communication channel can help.

What is the priority?
What has changed?
Is anything important enough to feed back?
What happens next?

That may be enough.

The learner should be able to describe some of the change

There is also a danger that progress becomes something adults discuss around the learner. Teacher says this. Tutor says that. Parent looks at the report.

But can the student answer:

What are you trying to improve?
What can you do now that was harder before?
What still catches you out?
What are you going to work on next?

If they can, tutoring goes beyond delivering support; it is helping the learner understand their own learning. That matters because feedback is most useful when learners have a clear next step and opportunities to act on it. A tutoring partnership should therefore produce not only better adult information, but better learner understanding.

Successful tutoring should eventually reduce the need for some tutoring

This may sound uncomfortable for a tutoring provider, but it matters.

If the objective of tutoring is to help the learner become more capable and independent, then some successful tutoring relationships should change over time.

Perhaps sessions become less frequent. The focus changes. The learner no longer needs help with the original problem. They move from individual tuition into a group. Or they stop altogether.

That is not necessarily lost business. It may show the original intervention worked.

Of course, a learner may continue tutoring for other reasons:

ongoing challenge;
exam preparation;
a new subject need;
or simply because they value additional teaching.

But tutoring should not create dependence simply to preserve itself.

A responsible provider should be willing to say: The original objective has been achieved. What does the learner need now?

Sometimes the answer will be: Something different.

Sometimes: Nothing.

“No improvement” is useful information too

Accountability should not mean pretending every intervention succeeds. Education does not work like that.

The learner may not respond to the tutoring approach. Attendance may be too inconsistent. The original diagnosis of the learning gap may have been wrong. The tutor–learner relationship may not work. The learner may need specialist support rather than more tuition. The programme may simply be ineffective.

Knowing this matters.

A responsible partnership should make it possible to say:

We tried this.
Here is what happened.
The evidence does not show the change we expected.
We therefore need to do something different.

That is a better outcome than continuing indefinitely because everyone can point to sessions delivered.

Evidence should not become a sales claim

This is especially important for tutoring providers. Progress evidence belongs to the learner.

It should first help:

the learner;
the family;
the tutor;
and, where appropriate, the school.

Providers should be cautious about turning individual progress into sweeping marketing claims. A learner improving after tutoring does not automatically prove that every future learner will improve in the same way.

EEF itself notes substantial variability within tutoring research and advises schools to consider context and professional judgement when applying evidence.

Responsible providers should make the same distinction. Evidence from this learner is not automatically proof of universal effectiveness.

Schools need to evaluate provision too

The principle matters particularly where schools fund tutoring. England's 2026–27 school census continues to record whether pupils have received school-funded tutoring. That helps show where tutoring is delivered, but the census does not capture the detailed educational impact of each learner's tuition.

That responsibility remains much closer to the school and provider.

So schools need to ask more than:

How many pupils received tutoring?
How many hours were delivered?
How much did it cost?

They also need questions such as:

Which learners received it and why?
What was the intended learning outcome?
Was the tuition sufficiently connected to classroom teaching?
What evidence changed?
Did the approach benefit all groups equally?
Should we continue, modify or stop it?

The more tutoring becomes part of ordinary educational provision, the more important these implementation questions become.

A simple evidence model

A school–tutor partnership does not need a complicated impact framework.

For an individual learner, five elements may be enough.

Starting point: What could the learner do before the support began?
Priority: What specifically are we trying to improve?
Support: What is the tutor actually doing to address it?
Evidence: What can the learner now do differently?
Decision: What should happen next?

That creates a simple chain: Starting point → Priority → Support → Evidence → Decision

Then the cycle repeats.

If the priority changes, change the tutoring.
If the learner becomes independent, reduce the support.
If the evidence is weak, investigate.
If progress is strong, move forward.

Simple, but much more meaningful than a record of hours.

Technology can make this easier

Technology has a clear role here. Not to manufacture progress or generate longer reports, but to make the evidence easier to connect.

A system could help bring together:

the learning objective;
relevant school context;
tutoring activity;
examples of learner work;
tutor observations;
assessment information;
and agreed next actions.

AI might help summarise patterns or surface changes, but the professional still needs to decide what those changes mean. Technology should reduce the administrative cost of seeing the learning journey; it should not replace educational judgement about whether that journey is going in the right direction.

The partnership should be able to answer one question

At the beginning of September, we argued that the strongest start to a school year is alignment, not activity.

One objective.
One starting point.
One next decision.

The same principle applies at the end of the month.

If school, family and tutor have been working together, we should now be able to ask: What changed?

Not:

How busy were we?
How many messages were sent?
How many sessions occurred?
How many resources were completed?

But:

What can this learner now do?
What remains difficult?
What evidence tells us that?
And what are we going to do next because of it?

That is the accountability that matters; not proving that tutoring happened.

Proving that somebody learned something - and using that evidence to decide what comes next.

If tutoring stopped tomorrow, what evidence would show what it had changed?

Sources and further reading

Education Endowment Foundation: One to one tuition
Evidence summary covering the impact of one-to-one tuition and practical implementation, including targeting learning gaps, linking tuition with classroom teaching, teacher–tutor communication and monitoring progress.
Education Endowment Foundation: Small group tuition
Evidence and implementation guidance for small-group support, including the importance of identifying specific needs, matching tuition to pupil understanding and linking additional support with classroom content.
Department for Education: Tutoring in education settings
Non-statutory guidance for schools, colleges and other 16–19 providers on planning and delivering tutoring.
Department for Education: School census 2026 to 2027: tutoring data
Current school-census guidance defining school-funded tutoring and requiring schools to identify pupils who have received it during the academic year.
Education Endowment Foundation: Developing Independent Learners: Feedback
2026 guidance emphasising clear learning goals, specific next steps and opportunities for learners to act on feedback rather than treating feedback as an endpoint.

We are discussing this article on LinkedIn: Tutoring happened. What changed?

For parents looking for trusted learning support: Find Tutors

Join the education movement: Teach. Create. Own.

For qualified teachers interested in joining TutorTech: Join as a Tutor

Read other Education Futures blogs:

Explore more education futures commentary, teacher-led perspectives and AI-in-education insights on the TutorTech blog.

⬅️ Back to Blog

A responsible school–tutor partnership should be able to prove what changed