How to plan a lesson: A step-by-step guide for teachers

Most teachers will tell you the same thing: A good lesson plan is not the document you hand to a mentor, it is the thinking you did before you walked into the room. The Word file with five neatly labelled sections matters far less than whether you can answer one question before the bell goes. What do I want them to know, do, or be able to explain by the end of this hour, and how will I check that it actually happened?

This guide walks through a planning process that tends to work for most teachers in most secondary classrooms. It is not a script. It is a sequence of decisions you make in roughly the same order, week after week, that gets faster the more you do it. ECTs often find a structured walkthrough useful because it gives a stable spine. Experienced teachers tend to compress the same steps into ten minutes on the train home.

What we are not going to do is romanticise it. Planning a lesson well is craft, not magic, and the research base on what helps students learn is reasonably clear. Rosenshine's Principles of Instruction and the EEF guidance reports on metacognition and assessment both point in the same direction: Clear outcomes, plenty of modelling, guided practice, frequent checks for understanding, and a meaningful chance to retrieve last week's material.

One of the most common planning traps, especially early in a career, is starting with an activity. You see a nice card sort on Twitter, or you remember the lab that worked beautifully last year, and you build the lesson around it. The activity becomes the lesson rather than the vehicle for learning something specific.

Flip the order. Begin by writing down, in one sentence, what students should be able to do by the end of the lesson that they could not do at the start. Keep it concrete and observable. "Understand photosynthesis" is vague. "Explain why the rate of photosynthesis plateaus at high light intensity, referencing the limiting factor" is something you can actually check.

A useful test: If a colleague walked into your room at the end of the lesson and asked three students the same question, would you be confident they could answer it? If not, the outcome is probably too broad or too abstract for one hour. Trim it down. One sharp learning outcome tends to beat three woolly ones for the same lesson slot.

Lesson plans live or die on what students bring through the door. New content lands on top of existing knowledge, and if the foundation is shaky, the new layer slides off. So before you plan the lesson itself, spend two minutes thinking about prior knowledge.

Three quick questions tend to be enough. What did they cover in the previous lesson on this topic, what did they cover in earlier years that this depends on, and what misconceptions are most likely to be in the room? For an A-level chemistry class learning about reaction kinetics, the prior knowledge might include collision theory from GCSE, rates of reaction practicals from Year 12, and a common misconception that catalysts get used up.

The planning move is to design something into the first ten minutes that surfaces this. A retrieval task on the previous lesson, a hinge question that diagnoses the misconception, or a quick whiteboard task that lets you scan the room. You are not pre-teaching. You are checking whether the ground is firm enough to build on, and adjusting if it is not.

Once you know the outcome and the starting point, the middle of the lesson tends to fall into a familiar shape. You explain the new idea, you show students what good looks like, and you give them progressively less support as they try it themselves. This is the I do, we do, you do sequence, and it lines up neatly with the research on explicit instruction.

The explanation matters more than teachers sometimes give it credit for. Plan the exact words you will use for the trickiest step. Decide where the analogy will be (and where it breaks down). Anticipate the two or three questions students will ask, and have an answer ready. Writing this part down, even as bullet points, often saves you from waffling on the day.

For modelling, the trick is to make your thinking visible. Do not just show the finished worked example on the board. Build it live, talking through the decisions, including the wrong turns. "I could try X here, but that would not work because Y, so I am going to try Z." Students who watch you reason aloud learn the process, not just the answer. This is especially powerful for subjects where the steps are not obvious from looking at the final product, like maths proofs, essay structure, or extended response answers in science.

Practice is where the learning actually consolidates, but the design of the practice matters as much as the time spent on it. The most common planning mistake here is making everything the same difficulty. Six near-identical questions teach students that they have mastered the technique. The first variation breaks the illusion.

A better pattern is to build a short practice sequence that ramps up. Two or three warm-up questions that look like the worked example. Two questions with one variable changed. One question that combines today's idea with something from last week. One question that asks them to apply the idea to an unfamiliar context. By the end, students should have met the same skill in three or four different shapes.

Leave room for guided practice between the modelling and the independent work. This is the we do phase. Do one example as a class with students contributing the next step. Do another one in pairs. Then release them to work alone, with a clear instruction about what to do if they get stuck. The drop in scaffolding should feel gradual, not abrupt.

If you only check what students know at the end of the lesson, you have already lost the opportunity to fix things. Checks for understanding need to happen at the points in the lesson where misunderstanding is most likely, which is usually right after the modelling and again partway through the practice.

Mini-whiteboards remain one of the simplest tools available. Pose a question, give students 30 seconds, ask them to hold up their answers. You get an instant read of the room and can decide whether to push on or re-explain. Hinge questions, where the wrong answers map to specific misconceptions, are particularly useful here. If you design one well, the pattern of incorrect answers tells you exactly what to address next.

Cold call is the other reliable workhorse. Ask the question first, pause, then name a student. The pause forces everyone to think. The naming spreads accountability across the room rather than letting the same three confident hands answer everything. Most teachers find this awkward at first. It gets easier quickly, and the increase in classroom thinking is noticeable within a week.

Lesson endings are often an afterthought, which is a shame because they are one of the highest-leverage parts of the hour. The brain consolidates what comes last more strongly than what comes in the middle. If the last thing students do is pack their bags while you shout the homework over the noise, that is what they will remember.

A good ending does three things. It pulls the key idea back into focus, it asks students to retrieve or apply it one more time, and it sets up the next lesson. An exit ticket works well: One question on a sticky note, answered in two minutes, handed in on the way out. You get a snapshot of who got it. Students get one more retrieval rep. The next lesson has a built-in starting point.

Keep the homework brief and connected. The research on homework in secondary settings is mixed, but the evidence suggests that short, well-connected homework tends to outperform long, disconnected homework. A 15-minute task that revisits today's lesson and feeds into next week's tends to land better than an hour of unrelated busywork.

If the planning process above feels heavy, that is partly because it is laid out for a first read. Once it becomes routine, the actual planning compresses into a short mental checklist you can run through in ten minutes per lesson. Here it is in shortened form.

A practical note for science and maths teachers in particular: A lot of the prep time in lesson planning goes into finding or writing exam-aligned practice questions and retrieval starters. If you have a question bank that maps to your specification, that bit gets faster. Cognito's teacher tools include exam-style questions, instant-feedback quizzes, and short videos tied to GCSE and A-level specs, which some teachers use as ready-made starters or homework. Not a magic bullet, but worth a look if you are spending an hour a week on retrieval questions you could otherwise reuse.