A lesson is a time budget
with an objective attached.

An objective students could recognise they had met: Write the objective as something a student can do at the end, not a topic you will cover. Compare the two: covering photosynthesis is unmeasurable, explaining why a plant in a dark cupboard dies is something a student can either do or not do by the bell.

A starter that earns its four minutes: The first activity should either retrieve prior learning the lesson depends on, or create the question the lesson answers. Put it on its own card with a hard time on it, because the starter is the single most common source of overrun.

Gradual release, with the minutes where the learning is: Lay out I do, we do, and you do as three cards and look at the split honestly. If modelling has twenty minutes and independent practice has eight, the lesson is a lecture with a worksheet stapled to it, and you can see that at a glance on a canvas.

A plenary that is not the bit you cut: Plan the last five minutes first and protect them. Consolidation and the check on what actually landed are what make the lesson stick, and they are always what gets sacrificed when the middle overruns.

Put the check in the middle, not at the end: A check at the end tells you what to do next lesson. A check at the twenty-five minute mark tells you what to do in the next twenty minutes. Pin the diagnostic question to a card in the middle of the row so it is part of the plan rather than an intention.

Plan the two branches: Draw two cards off the check: what happens if most of the class has it, and what happens if most of them do not. Both branches sit on the canvas, so the decision in the moment is a choice between two prepared paths rather than improvisation.

Differentiate in both directions: Pin the extension task for the student who finishes in four minutes and the scaffold, worked example, or sentence starter for the student who needs another pass. Both are attached to the practice card, so neither is something you remember you needed halfway through.

A first draft from the objective: Describe the objective, the year group, and the period length and the AI lays out a timed sequence with a starter, modelling, practice, a check, and a plenary. It is a starting position to disagree with, which is much faster than an empty template.

It reads the whole board: Because the AI has the whole canvas as context, you can ask it to suggest a better hook for the lesson that is actually on the board, to propose an extension task that matches the practice you wrote, or to list the misconceptions this topic usually produces.

Ask it what to cut: Most plans are over-stuffed by about a third. Ask the AI which activity is doing the least work for the objective, then decide for yourself. The judgement stays with you, because you are the one who knows the class.

Duplicate and adapt: Copy the board, change the example that did not land, swap the video, and adjust the timings you know were wrong. Adapting last year's lesson takes a fraction of the time of writing a new one, and the materials come with it.

Share a view-only link: Send a link to a colleague, a mentor, a trainee, or your head of department. They open the whole lesson in a browser with no account, which makes a conversation about pacing far more concrete than a paper plan.

Plan together: A department can build a lesson on the same canvas, so a shared scheme has one version of the lesson rather than four people's folders. Real-time collaboration is available on the Team plan.

Export for a folder or an observation: Export the board as a high-quality image or PDF when someone needs a flat copy, for an observation, a training file, or a printout to annotate during the lesson itself.