The Tutor Handbook · Volume 0187 · Series ID THB-0187
The Tutor Handbook: Complete Series Index
A good teaching routine can fail before anybody teaches it badly
A tuition programme finds a promising instructional routine. The research-informed guidance is sensible. Tutors agree with it in training. The demonstration lesson works. The routine is added to the handbook and everybody leaves the meeting believing implementation has begun.
Three weeks later, the practice is barely visible.
One explanation is poor fidelity: tutors did not do what they were asked to do. That explanation is sometimes correct. It is also dangerously convenient. The routine may require eight minutes of recording after every ninety-minute lesson, a second adult to observe learner responses, materials that take too long to prepare, a digital tool that is unreliable on the centre’s devices, or a sequence that assumes all three learners remain on one route. In the demonstration, those conditions were hidden by extra preparation, unusually cooperative learners or a coach who quietly carried part of the load.
The programme then blames people for failing to implement a practice that was never operationally realistic under ordinary conditions.
The Implementation-Feasibility Gate asks a different question before blame begins: can this evidence-informed routine actually fit the time, materials, skills, learner variation and support conditions of normal tutoring while preserving the part that makes the routine educationally useful?
Feasibility is not an excuse to weaken a practice until nothing remains. It is a test of whether the proposed practice and the actual operating environment can coexist.
Quick answer
A tutoring routine is feasible when ordinary tutors can deliver its load-bearing features to the intended learners, at the intended frequency and quality, with the time, materials, training, coordination and attention the programme can realistically sustain.
Before deciding that non-use is a tutor-fidelity problem, check whether the routine asks for more than the environment can supply. Identify the active ingredient. Separate it from optional surface features. Test the routine in a representative week rather than a showcase lesson. Measure the recurring burden, not only the setup effort. Watch whether the routine still works when a learner is absent, a group branches, school homework arrives, a device fails, or the tutor has another class immediately afterwards.
If the practice is valuable but infeasible, the programme has several options: simplify a non-essential surface feature, add capacity, narrow the situations in which the practice is required, improve materials or training, redesign the workflow, or decide not to adopt it.
What it should not do is preserve an impossible specification and call every deviation poor teaching.
Why this is a distinct implementation job
The Tutor Handbook already has an Implementation Fidelity Check. That owner asks whether a learning route was actually delivered with its intended active ingredients, dose, sequence, support conditions and learner participation before anyone declares the intervention ineffective.
The feasibility question comes earlier.
Imagine a programme asks every tutor to conduct a five-item retrieval review at the start of every session, record each learner’s error type, enter the result into a dashboard, generate one personalised follow-up item, and use that result to adjust the next lesson. The educational idea may be sound. Fidelity asks whether tutors performed the routine. Feasibility asks whether the full routine can realistically survive six back-to-back classes, three learners per group, existing reporting duties and the actual software available.
A practice can be feasible but implemented poorly. A practice can also be implemented faithfully during a short pilot while being infeasible as ordinary work. Those are different diagnoses and require different remedies.
The Instructional-Practice Consistency Gate owns another neighbouring question: how one teaching model remains coherent across different tutors without forcing identical scripts. Feasibility asks whether the model can fit ordinary operating conditions at all.
Apex guidance now makes feasibility explicit
The Australian Education Research Organisation published Staying on track: Monitoring implementation outcomes on 15 September 2026. It is a research-informed practice guide for school implementation teams, not a private-tuition trial. The guide distinguishes several implementation outcomes, including feasibility, acceptability, fidelity, reach and sustainability, and treats those outcomes as useful evidence for reflection and implementation decisions.
That distinction is useful because programmes often collapse everything into “Did staff follow the programme?” A practice can be acceptable but infeasible. It can be feasible but fail to reach the intended learners. It can be faithfully delivered for one month but unsustainable over a year.
The Education Endowment Foundation’s Implementation guidance, third edition published 24 April 2024, similarly treats implementation as a process shaped by context, people, behaviours and supporting structures rather than a one-off decision to adopt a promising idea.
The National Student Support Accelerator’s Toolkit for Tutoring Programs takes the same practical direction from a tutoring perspective. It repeatedly links programme design to tutor type, training, support, ratio, materials, session structure and delivery model. Those relationships matter because an instructional routine does not arrive in a vacuum. It lands inside an operating system.
These sources do not validate the exact gate proposed here. They make the neglected job visible.
Feasibility is about ordinary conditions, not best-case conditions
The easiest way to overestimate feasibility is to test a practice under privileged conditions.
A coach designs the routine, teaches it once, prepares the materials personally and has no class immediately before or after. The learners have good attendance. The technology works. Nobody asks for urgent school homework. The coach knows exactly what the routine is meant to achieve because they designed it.
The lesson works.
That is evidence that the routine is possible. It is not yet evidence that the routine is feasible.
Feasibility asks what happens when the practice is delivered by an ordinary competent tutor on an ordinary week. The tutor has the normal preparation window, the normal class transitions, the normal range of learners, the normal devices and the normal amount of supervisory access.
This is not an argument for lowering standards to match weak systems. It is an argument for telling the truth about what the system currently supports.
If the routine needs more capacity, say so. Then decide whether the educational value justifies creating that capacity.
The load-bearing feature comes first
A feasibility review becomes dangerous if the first response to operational difficulty is to simplify the practice. Simplification can remove the active ingredient.
Suppose a feedback routine has three essential parts: identify the gap, give the learner an opportunity to act on the feedback, and later check whether the improvement survives a fresh attempt. Tutors report that the routine takes too long. A manager responds by keeping the written comment but removing the reattempt.
The routine is now easier. It may no longer be the same educational routine.
The existing Feedback-Action Loop makes this problem visible. Feedback without learner action can become information that never changes performance.
Before redesigning for feasibility, state the active ingredient in plain language. What must remain true for the practice to do its intended job? Which features are support structures, recording conveniences or preferred formats rather than mechanisms?
A routine becomes feasible by changing what can change, not by quietly removing what makes it work.
A feasibility map for tutoring
A programme does not need an elaborate scoring model. It needs to examine the recurring constraints that can break the practice.
Time. How many minutes does the practice consume during the lesson, before the lesson and after the lesson? Does it displace independent practice, feedback, preparation or recovery time?
Tutor attention. Does the routine require the tutor to watch all three learners simultaneously while also recording detailed data? Does documentation make the tutor miss live reasoning?
Materials. Are the required questions, examples, texts, manipulatives or digital tools actually available at the right level and in the right format?
Tutor knowledge and skill. Does the routine depend on diagnostic or facilitation moves that tutors have not yet learned to use in live lessons?
Group architecture. Does the routine assume all learners are on one route when real groups frequently branch?
Technology. Does implementation depend on devices, internet access, accounts or integrations that fail often enough to distort ordinary use?
Coordination. Does the routine require information from school, parents or another tutor that arrives too late or inconsistently?
Data burden. Does the programme collect information that nobody later uses to make a decision?
Transition cost. How much time does it take to switch into and out of the routine? A five-minute activity can cost eight minutes if setup and reset are ignored.
Recovery. What happens when the routine fails? Is there a workable offline, low-tech or shorter path that preserves the learning job?
These are not excuses. They are design variables.
Composite case: the progress-monitoring routine that consumed the lesson
The following case is fictional and constructed for teaching.
A tuition programme introduces a detailed weekly progress-monitoring routine. Each learner completes six short items. The tutor codes the error type, enters results into a spreadsheet, writes a brief interpretation and generates a personalised follow-up task. The programme expects the routine to take fifteen minutes.
During training, it does.
In live three-learner tuition, the routine regularly takes thirty minutes. One learner finishes quickly and waits. Another needs the instructions clarified. A third has a legitimate access support that changes how the task is presented. The tutor then spends eight minutes after class entering data while the next group is already arriving.
After a month, tutors begin shortening the routine. Some skip the error codes. Some enter results later from memory. Some give the same follow-up task to all three learners. Leaders see reduced fidelity.
The immediate temptation is to retrain the tutors.
A feasibility review finds a different story. The six-item set is longer than necessary for the decisions being made. The spreadsheet duplicates notes already kept elsewhere. The personalised follow-up can be chosen from three prepared branches rather than generated from scratch. The programme redesigns the surface workflow while preserving the active ingredient: obtain a small sample, interpret the result and let the interpretation influence the next teaching move.
The revised routine may now be feasible. The original implementation problem was partly a design problem.
Feasibility should be tested against a representative week
One lesson rarely reveals recurring cost.
A routine that adds five minutes to Monday’s lesson may be harmless. The same routine repeated across twenty groups may create hours of after-class work. A digital tool that works on the programme lead’s laptop may fail on the actual devices tutors use. A sophisticated branching protocol may be manageable in a stable group and collapse during a week when several learners return from absence.
A representative-week test asks tutors to run the routine under normal operating pressure and preserve simple evidence about what it displaced.
The programme can ask which part took longer than expected, which part tutors skipped first when time tightened, which learner conditions made the routine harder to use, which information was collected but never used, which adaptation tutors invented independently, whether the routine still created the intended learner opportunity, and what happened in the next lesson because of the information collected.
The point is not to make tutors prove the programme wrong. It is to find the true cost before the routine becomes compulsory.
Composite case: a discussion routine that depends on invisible coaching
This case is fictional.
A coach demonstrates a three-learner discussion routine. Each learner first thinks privately, one explains, a second compares, and a third challenges or extends. The routine is excellent in demonstration.
Tutors later struggle. Their discussions become shallow. Leaders conclude that tutors do not understand dialogic teaching.
Observation reveals that the coach was making several expert decisions that were never written into the routine. She chose problems with multiple plausible approaches. She knew which learner to call first. She noticed when a comparison would be meaningful and when it would merely repeat the answer. She could revoice without upgrading weak reasoning. Ordinary tutors had received the visible routine but not the professional judgement underneath it.
The feasibility problem is partly a training problem. The routine asks tutors to perform a more advanced facilitation function than the programme has developed.
The answer may be to add rehearsal, case comparison and coaching before expecting routine use. Alternatively, the programme may narrow the routine to lesson types where tutors can use it reliably while skill develops.
Calling the original failure “resistance” would miss the mechanism.
Feasibility and acceptability are not the same
Tutors can dislike a feasible routine. They can also like an infeasible one.
A new note-taking method may be quick and easy but feel pointless because tutors do not understand how the information will be used. A rich coaching routine may be valued highly while demanding more observation time than the programme can provide.
These are different implementation outcomes.
Acceptability asks whether people regard the practice as appropriate, reasonable or worthwhile. Feasibility asks whether it can actually be carried out within the available constraints. Both matter.
If tutors dislike a routine because its educational purpose is unclear, better explanation may help. If tutors dislike it because it adds forty minutes of work after every evening class, communication alone will not solve the problem.
Leaders should not turn every practical objection into an attitude problem.
The resource fantasy
A common planning mistake is to design for resources that technically exist but are not operationally available.
The centre owns tablets, but only four are charged. The question bank contains thousands of items, but tutors cannot search it quickly enough during preparation. A coach is assigned to support tutors, but also teaches a full load and cannot observe when needed. Parent information exists, but arrives through different chat threads and cannot be found reliably.
On paper, the resource exists. In practice, access friction makes it unavailable at the moment the routine needs it.
Feasibility therefore includes usable access, not inventory.
A practical test is: can a tutor obtain the resource, understand it and use it correctly inside the real workflow without extraordinary effort?
If the answer depends on the most organised tutor in the programme, the system is not yet feasible for ordinary implementation.
Do not confuse a pilot with ordinary implementation
Pilots receive attention. Leaders watch closely. Problems are fixed quickly. Participants know they are testing something. The practice often receives more coaching than it will ever receive at scale.
This is useful. It is also a different condition.
Before adoption, the programme should identify which pilot supports will disappear. If the pilot used a weekly coaching meeting, who carries the routine after that meeting stops? If the programme lead prepared all materials during the pilot, who prepares them later? If pilot learners were selected because attendance was reliable, how will the routine work for the full population?
A practice that succeeds only while surrounded by exceptional support may still be valuable. The programme simply has to decide whether that support should become part of the permanent model.
Feasibility can differ by learner group
A routine may be feasible for one part of the programme and not another.
A brief oral review may fit a fluent Secondary group but take much longer for learners who need additional processing time. A digital annotation routine may work for learners with reliable devices but create access barriers elsewhere. A peer-comparison activity may work in a stable group and become confusing in a newly formed group whose members are on different curriculum routes.
The correct response is not always universal adaptation. Sometimes the routine should have declared eligibility conditions.
“Use this routine when learners share the target and can generate a private first response” is more honest than “all tutors must use this routine every lesson”.
This protects the programme from mistaking selective appropriateness for non-compliance.
Feasibility should preserve accessibility
A practice is not feasible merely because most learners can use it.
If the routine depends on reading tiny text, rapid oral response, a specific device interaction or an unnecessary motor demand, access barriers may prevent the intended learner from participating in the target task. The programme should preserve legitimate access supports and distinguish them from instructional help that supplies the answer.
The existing Access-Support Boundary owns that distinction.
The feasibility question is operational: can the practice be delivered under legitimate access conditions without losing the intended learning job?
If not, redesign is required.
Tutor capacity is part of feasibility but not the whole of it
The Tutor Capacity Boundary asks whether a tutor has enough preparation, observation, follow-up and recovery capacity before taking on more work.
Implementation feasibility is wider.
A tutor may have enough time but lack usable materials. They may have good materials but insufficient training. They may have the skill but be placed in a group architecture that makes the routine impractical. They may have all of those and still face a reporting workflow that duplicates data.
Feasibility therefore looks at the tutor inside the operating system.
This avoids the familiar management error of treating every implementation problem as an individual productivity problem.
A minimum viable routine must still do the educational job
Programmes often use the phrase “minimum viable” to mean “shortest possible”. That is not enough.
The minimum viable version should be the smallest routine that still preserves the mechanism. If the practice is retrieval followed by corrective feedback, removing the feedback may change the routine. If the practice is learner use of feedback, removing the reattempt changes the job. If the practice is representative progress monitoring, shrinking the sample until it no longer covers the target creates false efficiency.
A useful redesign question is: what is the smallest version that would still make us recognise the same educational practice?
Everything below that line is not simplification. It is substitution.
Composite case: the AI question generator that saves no time
This case is fictional.
Tutors begin using a generative AI tool to create personalised practice. The promise is efficiency. Tutors can produce five tailored questions in seconds.
In reality, the questions require verification. Some contain ambiguous wording. A few answers are wrong. Difficulty varies unpredictably. Tutors spend time editing formatting and checking whether the questions match school methods.
The programme counts generation time and concludes the tool saves preparation. Tutors experience the opposite.
The AI Material Verification Gate already establishes that learner-facing material must be verified. The feasibility question is whether the entire verified workflow is actually cheaper or better than the alternative.
If not, the tool may still be useful for unusual cases rather than routine generation. Narrower use can be the honest feasible design.
Five responses to an infeasible practice
When a valuable routine fails the feasibility gate, programmes have more choices than enforce or abandon.
Simplify the surface. Remove duplicate recording, unnecessary formatting or redundant steps while preserving the active ingredient.
Add capacity. Provide preparation time, coaching, better materials, functioning technology or clearer ownership.
Narrow eligibility. Use the routine only where the learner configuration and lesson purpose justify it.
Redesign the workflow. Move preparation earlier, automate a safe administrative step, create shared materials or merge duplicate records.
Reject or defer. Sometimes the programme is not ready. A promising practice can remain a future option rather than becoming a badly implemented present requirement.
Each response is more informative than blaming tutors by default.
Feasibility should be rechecked after the novelty period
A new routine can feel difficult because it is unfamiliar. Tutors need more attention to remember the steps. Preparation takes longer. Learners need the routine explained.
That does not prove long-term infeasibility.
The programme should distinguish learning cost from steady-state cost. After tutors have used the routine enough to become fluent, does the burden fall to a sustainable level? Do learners understand the structure? Do materials become reusable?
The opposite can also happen. A practice feels easy during the first month because enthusiasm and extra support are high. Burden appears later when records accumulate, staff change, materials need maintenance or the coach can no longer provide constant help.
A delayed feasibility check is therefore essential.
Sustainability is the next question, not a synonym
AERO’s implementation framework separates sustainability from feasibility for a reason.
Feasibility asks whether the practice can be carried out under current operating conditions. Sustainability asks whether it can continue over time while the system changes.
A practice can be feasible this term because one expert tutor maintains the materials. It may not be sustainable if nobody else can update them. It can be feasible while enrolment is low and collapse when groups fill. It can be feasible while a temporary project funds coaching and become unsustainable when that support ends.
The Tutor Handbook does not need to turn these concepts into bureaucracy. It needs tutors and programme leaders to stop using “works” as a single undifferentiated word.
What parents should experience
Parents do not need an implementation-science report.
They should experience a programme whose routines actually fit the service being promised. If the centre says tutors monitor progress, the monitoring should influence teaching rather than exist only as paperwork. If the centre says lessons are personalised, tutors should have enough time and usable information to make real adaptations. If the centre says feedback is acted upon, learners should receive a chance to use it.
Feasibility is invisible when it is done well. The programme simply works without requiring tutors to hide impossible workloads or families to compensate for operational gaps.
Failure modes
The blame-the-tutor failure. Leaders see low use and assume weak commitment before checking whether the routine fits ordinary work.
The pilot-privilege failure. A practice succeeds with extra preparation, coaching and selected learners, then is treated as proven feasible at scale.
The resource-fantasy failure. Tools and support technically exist but cannot be accessed at the moment the routine needs them.
The paperwork-fidelity failure. Tutors complete forms that make implementation visible while the learner-facing mechanism receives less attention.
The active-ingredient amputation. A routine is simplified until the feature that creates learning has disappeared.
The expert-dependence failure. The practice works only when one unusually experienced tutor or programme lead is present.
The universal-use failure. A practice that is feasible and useful for some lesson conditions is mandated everywhere.
The novelty confusion. Short-term learning cost is mistaken for permanent infeasibility—or initial enthusiasm is mistaken for long-term sustainability.
The hidden access failure. The routine is feasible for the easiest learners but systematically difficult under legitimate accessibility conditions.
The burden-export failure. Tutor workload is reduced by pushing extra coordination or teaching labour onto parents or learners.
A practical feasibility review
Before adoption, write one sentence stating the active ingredient. Then estimate the full recurring workflow: preparation, delivery, recording, follow-up and maintenance. Test the routine with ordinary tutors and ordinary groups. Note where it breaks. Ask whether the break reflects missing skill, missing resource, poor workflow, inappropriate eligibility or a weak routine.
Redesign only the parts that can change safely.
Then test again after the novelty period. If the routine now fits and the active ingredient remains intact, the programme has evidence of feasibility. If not, the honest decision may be to narrow, defer or reject it.
This is not anti-innovation. It is how innovation becomes operational rather than ceremonial.
Sources and evidence limits
AERO’s Staying on track: Monitoring implementation outcomes was published and updated on 15 September 2026. It is a research-informed practice guide for school implementation teams and explicitly discusses feasibility, acceptability, fidelity, reach and sustainability as implementation outcomes. This article adapts that vocabulary to tutoring; it does not claim that AERO validated this exact feasibility gate.
The Education Endowment Foundation’s Implementation guidance, third edition published 24 April 2024, is implementation guidance for education settings rather than a private-tuition trial. It supports attention to process, context, people and supporting structures.
The National Student Support Accelerator’s Toolkit for Tutoring Programs is current tutoring-programme guidance that links instructional design to tutor selection, training, support, ratio, content and delivery. It is used here as a high-authority design source, not as proof of one universal operating model.
The exact feasibility thresholds proposed in this article are professional decision rules. They should be tested against local tutoring evidence and revised when ordinary conditions change.
The end state
A tutoring programme should not ask tutors to perform educational theatre.
If a routine matters, build the conditions that let ordinary tutors use it well. If the conditions cannot be built, narrow or redesign the routine. If the routine survives only when its active ingredient is removed, call the new practice something else.
The strongest implementation culture does not choose between fidelity and practicality. It identifies what must remain true, then designs a system in which remaining true is actually possible.
That is the Implementation-Feasibility Gate.
Before asking, “Why didn’t the tutor follow the routine?” ask the more revealing question: “Could a competent tutor really have followed it here, this often, for these learners, without sacrificing the very learning the routine was meant to improve?”
Feasibility debt accumulates when workarounds become invisible
A routine can appear feasible because tutors quietly subsidise it.
They stay fifteen minutes after class. They build their own spreadsheets. They keep duplicate notes because the official record is too slow. They prepare questions on weekends. They remember learner states instead of using the intended continuity system. They ask a colleague informally because the formal coaching route takes too long.
The programme sees successful implementation and concludes the design fits.
What it is actually seeing is feasibility debt: recurring work that has been pushed outside the declared model. The debt may be carried for months by conscientious tutors before it becomes visible through burnout, inconsistent records, staff turnover or sudden collapse when the person who held the workaround leaves.
A feasibility review should therefore ask not only, “Did the routine happen?” but, “What invisible labour made it happen?” That question is especially important before scaling a practice from one tutor to several. If the routine depends on one person’s private template, memory or unpaid preparation, the programme has not yet designed a transferable system.
This does not mean every extra minute is unacceptable. Professional work always contains judgement and preparation that cannot be reduced to a mechanical schedule. The issue is whether the recurring burden is known, proportionate and supported. Hidden subsidy creates false confidence because leaders design the next stage using costs that only tutors can see.
A practical way to surface the debt is to ask tutors what they would have to stop doing if the routine doubled in frequency tomorrow. Their answer reveals the displaced work. If the only way to scale progress monitoring is to reduce feedback quality, if the only way to use a new AI workflow is to stop verifying questions carefully, or if the only way to maintain documentation is to stop reviewing learner work before class, the routine is competing with another active ingredient of good tutoring.
The programme then has a real design choice rather than an imaginary free improvement.
Scale changes feasibility even when the routine stays identical
A practice that fits one tutor can become infeasible across a programme because shared resources become bottlenecks.
One coach can support two tutors intensively. The same coach may not support twelve. One material designer can prepare a specialised question set for three groups. The same workflow may fail when every level and subject requests weekly customisation. A single shared device can support occasional use and become a queue when the routine becomes standard.
Scaling therefore changes the denominator of feasibility. The active ingredient may remain identical while the support ratio, maintenance burden and coordination network change around it.
Before programme-wide adoption, leaders should identify which resources scale with the number of tutors or learners and which do not. Some supports duplicate cheaply: a vetted template, a recorded demonstration, a reusable task bank. Others require continuing human attention: observation, case consultation, material checking, individual parent coordination.
This distinction protects the programme from assuming that a successful small pilot is merely a smaller version of full implementation. Sometimes scale requires a different support architecture.
The honest question is not, “Can this work?” It is, “Can this work at the volume we are about to require, while preserving the rest of the tutoring system?”
