A lesson can be brilliantly delivered and still fail the person it was meant to reach. The teacher makes a difficult idea clear; the examples are elegant; the class nods. Twenty-four hours later, a student faces a slightly different problem and cannot begin. The presentation was successful. The learning was not yet secure.
That gap—between what an adult has explained and what a learner can independently do—is the provocation at the heart of Stop Teaching. Start Coaching., a Learning Coach book whose title sounds almost hostile to the profession it seeks to reimagine. But its target is not teaching itself. It is the habit of treating the completion of teaching as sufficient evidence that learning has occurred.
The book's published architecture is ambitious: five parts, 22 chapters, and a journey from the history of mass schooling through the study skills we demand but often leave implicit, the pressures on teachers and parents, the proposed role of a learning coach, and the arrival of generative artificial intelligence. The question beneath its slogan is more serious than the slogan: What is a school for when explanations can be delivered everywhere, but understanding cannot be assumed anywhere?
The title picks a fight that the book itself complicates
“Stop teaching” is an excellent headline and a questionable literal policy. Children do not acquire algebraic procedures, fluent reading, scientific vocabulary or historical context simply because an adult asks stimulating questions. Novices benefit from carefully chosen examples, explicit explanations, models, practice and corrective feedback. A school that stopped teaching in the ordinary sense would quickly betray its students.
The official description of Stop Teaching. Start Coaching. explicitly acknowledges this. The book is not framed as a rejection of clear instruction, particularly for beginners. Its opposition is to an educational endpoint: if the lesson has been delivered, the job is considered finished. By shifting the emphasis from the educator's performance to the learner's developing capability, the title becomes a provocation about responsibility rather than a proposal to abolish pedagogy.
This is an important distinction because the vocabulary of educational reform is full of false choices: knowledge or skills, freedom or structure, explanation or discovery, rigor or care. A genuinely useful learning-coaching approach should be judged not by which side it chooses, but by whether it knows when a learner needs clear instruction, when the learner needs practice, and when the adult ought to withdraw support.
There is a credible research conversation here. The Education Endowment Foundation's second-edition guidance on metacognition and self-regulated learning recommends explicit teaching and modelling of strategies, embedded within subject learning, followed by scaffolding and increasing independence. Those are not competing activities. They describe a progression.
What mass schooling achieved—and what its routine can conceal
The first part of the book, titled We Invented Teaching, is a smart place to begin. School is not a natural law; it is an institution with a history. Age-grouped classes, schedules, curricula and examinations arose partly because societies needed ways to make education available at a scale that private tutoring never could. A credible critique of the classroom has to recognize that expansion of access before declaring the format obsolete.
Yet a system built to reach many learners must simplify something. The period ends at a set time. The syllabus advances. The teacher must make consequential decisions for a group. Students who need a slower explanation, a missing prerequisite or a different practice schedule do not always fit the timetable. This is not proof that teachers are indifferent. It is often proof that the institution asks one professional to solve several incompatible logistical problems at once.
The book's published chapter list includes The Miracle of Mass Schooling and The Mistake We No Longer See. That pairing signals the argument at its best: education can be historically successful and pedagogically incomplete at the same time. An institution that widened participation should be preserved where it works and redesigned where its inherited assumptions make learning harder to see.
For the United States, that tension feels familiar even without endorsing one national diagnosis. School leaders must reconcile grade-level standards, diverse starting points, limited planning time, accountability demands and families' expectations. The answer cannot be a romantic promise that every teacher will instantly become a personal coach to every child. It has to begin with routines that make understanding more visible within the constraints teachers actually face.
The missing subject: how to learn
The most compelling charge in the book's public account is that schools repeatedly demand abilities they have not always taught explicitly. Concentrate. Organize. Revise. Persist. Study. Think independently. These imperatives are delivered as if they were straightforward instructions rather than complex activities that depend on prior knowledge, strategy, context and practice.
The book's prologue, quoted on its publisher's page, compresses this contradiction into one sentence: “We demand the result without teaching the procedure.”
It is a useful accusation, not a universal verdict. Many excellent teachers already teach these procedures. But the distinction matters. Telling a child to “study harder” is not the same as showing the child how to retrieve information from memory, identify an error, schedule review, or check whether a solution really follows.
Consider two students preparing for a science assessment. Both spend an hour with their notes. One repeatedly rereads a familiar page and experiences the comfort of recognition. The other tries to reconstruct an explanation without looking, discovers missing links, revisits those sections, then tests the explanation again after a delay. They have both been busy. Their activities differ in what they reveal about memory.
The US Institute of Education Sciences' What Works Clearinghouse guide on organizing instruction and study recommends such practices as spacing learning, working between examples and problems, using retrieval, and prompting deeper explanatory thinking. It also describes the difficulty learners can have in judging what they actually know. These findings support the importance of teaching study methods rather than assuming students will discover effective methods by chance.
The National Academies' How People Learn II places metacognition, executive function and self-regulation within a larger account of learning. Students need ways to monitor their own thinking, regulate attention and behavior, and modify strategies when conditions change. None of these abilities is a single “learning style,” and none eliminates the need for content knowledge. They are part of the work required to use knowledge well.
An error is information before it is a judgment
Imagine a student who cannot solve a word problem. The first adult interpretation might be lack of effort. The second might be a missing multiplication fact. The third might be difficulty translating language into a mathematical relationship. Those explanations demand different interventions.
A coaching-oriented response begins by collecting evidence. “Show me where you stopped knowing what to do.” “Which part of the example still makes sense?” “What would you try first if I were not sitting here?” The point is not to turn every lesson into a lengthy conversation. It is to avoid applying the same remedy to different causes.
There is also an ethical advantage in this habit of observation. When students repeatedly hear that difficulty proves laziness or incapacity, they may receive a moral judgment where a teachable problem exists. The reverse mistake is possible too: not every unsuccessful attempt reflects a missing strategy. Adequate instruction, accessibility, material resources, mental well-being and the demands of a task all matter. Careful observation must remain open to more than one explanation.
Coaching cannot be another name for withholding instruction
Here the book's central proposal meets its most important challenge. Learning coaching sounds humane and modern, but without boundaries it can become a vague instruction to “facilitate” while students struggle alone. The burden of discovery can quietly fall on those who have the least prior knowledge or outside support.
A strong educator has a repertoire, not a single preferred posture. At the start of a difficult skill, a worked example may be the most compassionate intervention. Later, guided practice allows the teacher to detect misunderstandings. With greater mastery, the adult can ask the learner to plan, explain, compare solutions, correct mistakes and work independently.
The EEF's guidance makes the logic explicit: teachers should model thinking, provide temporary support and gradually remove scaffolding as pupils become more capable. Independence is an outcome that must be built; it should not be confused with leaving learners unsupported.
The title Stop Teaching. Start Coaching. therefore succeeds only if it is read as a challenge to change the function of teaching, not to erase teaching. Its strongest version is less theatrical and more demanding: explain precisely, assess honestly, design practice intelligently, and transfer responsibility when the evidence shows the learner is ready.
What would count as progress?
A completed worksheet is weak evidence of autonomy when the teacher supplied every prompt. A polished essay is weak evidence when the student cannot explain a central claim. A confident declaration of understanding is weak evidence when retrieval fails the next day.
More informative questions are possible. Can the learner begin a comparable task without a cue? Explain why a solution works? Recognize a likely error? Select a strategy instead of guessing? Return to a problem after feedback? Apply an idea in a new but related context?
These are not flawless measures, nor do they replace formal assessment. They direct attention toward something worth measuring: what remains when the adult's help is no longer available.
The exhausted adult is part of the equation
The book's third part, Adults Running on Empty, moves the argument from an individual learner to the adults and institutions surrounding that learner. This is essential. It would be perverse to diagnose an organizational problem and then prescribe another obligation to an already overextended teacher.
A classroom approach requiring constant individualized conversation, without time, training or staffing, is not a reform; it is a transfer of institutional cost onto teachers' unpaid attention. The practical questions are unavoidable. How much planning time does meaningful diagnosis take? Which routines can be designed for the whole class? What assessment information is useful enough to change instruction? When can peers help each other, and when does a teacher's expertise remain indispensable?
The OECD's TALIS 2024 analysis of teacher leadership and autonomy reports associations between greater instructional autonomy and teachers' self-reported ability to adapt explanations and attend to students' prior knowledge and needs. These are associations, not proof that autonomy causes better learning. Still, they raise a structural question: can a school ask teachers to respond flexibly to learners while restricting their professional judgment about how to teach?
The term learning coach must not become a managerial excuse to demand more emotional labor without giving teachers workable conditions. If the aim is better diagnosis and more independence, school systems have to create space for the professional decisions those aims require.
AI has made the book's question harder to evade
The book's fifth part, AI Changes the Rules, gives the older teaching-versus-learning debate a new urgency. A student can now obtain an explanation, a summary, a worked solution or an essay draft with very little friction. This can improve access to useful help. It can also produce a dangerous illusion: the work appears to have been completed, so the learning appears to have happened.
A clean answer on a screen is not evidence that the student understands its premises. An AI-generated explanation may be clear while being mistaken. A student may submit a convincing argument yet remain unable to defend it. Conversely, a learner who asks an AI assistant to challenge a hypothesis, compare approaches and explain an error might use the tool in ways that support meaningful study.
The difference lies less in the mere presence of AI than in the cognitive role assigned to the learner. Who chooses the question? Who evaluates the answer? Who notices contradictions? Who can reproduce the reasoning with the tool closed?
UNESCO's guidance for generative AI in education and research argues for human-centered, age-appropriate and pedagogically justified use, with human agency preserved. Its approach does not treat the technology as an automatic solution to educational problems. That restraint is a useful test of any coaching philosophy in the AI era.
The implication is not that memorization has become obsolete. In many domains, knowledge held in memory is what makes it possible to judge an answer, detect nonsense and ask a fruitful next question. Nor does every learner need to complete every task without assistance. Rather, schools need to distinguish supported performance from independent competence, and to assess both intentionally.
A practical measure might be simple: after using AI to explore a problem, can the learner explain the answer, identify its limits and solve a related problem without the same assistance? If not, the tool may have improved production more than understanding.
Where the book's argument deserves resistance
Three objections should accompany any serious engagement with its premise.
First, the teacher–coach contrast can become a caricature. Excellent teachers have always observed misunderstanding, adjusted instruction and cultivated independence. The profession does not need to be told that all previous teaching was merely the transmission of facts. The book's value is in sharpening the distinction between activity and learning, not in claiming that one branded role invented responsiveness.
Second, independence is not equally easy to acquire. Families differ in time, space, money, familiarity with school expectations and access to specialist help. If schools expect children to manage their own learning before teaching them how, they may reward the students already receiving effective support elsewhere. A fair coaching model must include explicit strategy instruction and accessible support as an entitlement, rather than a premium service available to a few.
Third, a compelling educational philosophy is not the same thing as an evaluated intervention. Evidence for retrieval practice, formative assessment, metacognitive strategy instruction or scaffolded teaching cannot be treated as proof that every commercial program called “learning coaching” works. The book offers an argument and a proposed professional orientation. Specific implementations should still be evaluated for feasibility, outcomes, equity and cost.
These objections do not refute the book's thesis. They keep it accountable. A reform that begins by asking educators to attend more closely to learners must be willing to submit its own methods to the same scrutiny.
What a teacher could change on Monday
The most useful response to a provocative book is neither applause nor dismissal. It is a design experiment small enough to test.
Suppose a teacher is introducing a new mathematical method. The lesson can begin with a short retrieval check that identifies missing prerequisites. The teacher then models a representative example, thinking aloud about why each step is chosen. Students try a similar task with limited prompts, and the teacher inspects where attempts break down. Before moving on, students explain a decision they made and identify an error they might watch for next time. A later review checks what remains after the first exposure.
Now add one explicitly coached habit: each student records a question they could ask themselves when stuck—“What do I know?”, “Which example is structurally similar?”, or “How can I check this answer?” At first the teacher supplies these questions. Over time students are expected to choose among them.
This is not a miracle program, and it is not an instruction-free classroom. It is a practical translation of the book's central question into observable decisions: diagnose, model, practice, check, adapt, and gradually release responsibility.
For school leaders, a parallel experiment might begin by asking teams to inspect a particular transition: after the explanation and before independent work, what evidence determines whether students are ready to proceed? The resulting discussion is more useful than asking everyone to adopt a new label.
Who should read Stop Teaching. Start Coaching.?
Teachers weary of reforms that promise to replace expertise with enthusiasm may find the title irritating. They may also find in the book's published argument a defense of something they know from experience: finishing a lesson and securing learning are different accomplishments.
School leaders can use its five-part structure as a provocation about timetable design, assessment, professional discretion and what schools consider success. Parents may recognize the difference between demanding effort and teaching a child what useful effort looks like. Tutors and learning professionals will find a pointed challenge to measure their work by decreasing dependence rather than increasing the number of paid explanations.
Readers seeking a tightly tested experimental comparison of coaching programs should be cautious: the publisher describes this book primarily as a research-informed educational rethink, not a clinical or experimental trial report. Its publicly available material supports a serious discussion of the thesis and structure, but not a definitive appraisal of every chapter's evidence or argument.
The real question comes after the lesson
The temptation is to read Stop Teaching. Start Coaching. as a slogan about replacing one profession with another. Its more enduring invitation is to ask a question that survives every fashion in education: What can this learner now understand, remember, question and do that they could not do before—and what can they do without us?
Answering it well requires substantial knowledge, expert instruction, careful assessment, time, relationship and opportunities to struggle productively with support. It is not anti-teaching. It is a higher standard for what teaching is supposed to achieve.
The lesson ends when the clock says it ends. Learning has a different timetable.
Further reading and sources
Stop Teaching. Start Coaching. — official book overview, five-part outline, 22 chapters and attributed prologue excerpts, Learning Coach Editions. Primary publisher description; not a substitute for access to the full manuscript.
Metacognition and Self-Regulated Learning, second edition, Education Endowment Foundation, 2025.
Organizing Instruction and Study to Improve Student Learning, Institute of Education Sciences / What Works Clearinghouse, 2007.
How People Learn II, chapter 4: Processes That Support Learning, National Academies of Sciences, Engineering, and Medicine, 2018.
Teacher leadership and autonomy: Results from TALIS 2024, OECD.
Guidance for Generative AI in Education and Research, UNESCO, 2023; page updated 2026.
From Teacher to Learning Coach: What Changes in the AI Era?, Learning Coach — related article on classroom practice in the AI era.
