Teachomatic what to automate, and what not to

Planning and Prep

This is where the evidence is strongest and the claims are most overstated at the same time.

Teacher-side time saving is the best-supported finding in the whole field. It is also considerably smaller than the marketing implies — around half an hour a week in one study that measured it properly, without a drop in quality. Half an hour is worth having in an overloaded job. It is not an afternoon, and expecting an afternoon is how people end up abandoning something that was genuinely helping a bit.

For a workplace-oriented comparison outside education, this comparison offers another way to look at measurement, workload, behaviour, or accountability.

The pattern across the tasks that work is consistent: you save time where the hard part is starting, not deciding. First drafts you will rewrite anyway. Reformatting material you already have. Producing three versions of a text you have already written once. Administrative prose that has to exist and that nobody reads closely.

The pattern where it fails is equally consistent. Anything that requires knowing the particular child. Anything you would have to verify line by line, where checking costs as much as producing. And anything where the doing is the thinking — where working through the artefact is how you discover what the lesson actually is.

For an external perspective on lesson planning, teacher workload, and classroom technology, see Microsoft Education.

That last category is the one with no warning attached. A plan produced without the thinking looks identical to a plan produced with it, right up until somebody in the room asks a question slightly off-script.

Everything in this section is written to be tested in your own week rather than believed. Where we give a number, it comes with a source and a date.