Teachers will not keep them,
and the ones who do distort them.
The obvious way to measure workload is to ask teachers to log their hours, and it fails in every school that tries it. Most teachers stop within a fortnight because the logging is itself workload. The few who persist are the conscientious ones, whose weeks are not typical, and the act of logging changes what they do. A term of timesheets produces a spreadsheet nobody trusts and a staff room that resents it. The alternative is cruder and more honest: ask each teacher, at two points a term apart, to estimate their typical week in three blocks, and compare the estimates. Self-report is imprecise, but it is imprecise in the same direction both times, and the difference between the two is a number a school can act on.
It is also the only number that matters to your decision. A vendor's percentage — including the one we quote, with its caveats, in the previous article — tells you what happened somewhere else. Six of your own teachers, asked the same five questions twice, tell you what happened here.
Three numbers, one hope,
one honesty check.
Keep the wording exactly the same both times; a changed question produces a changed answer. Ask for hours, not for "a lot" or a scale of one to five, because hours can be added and compared. Separate preparation, marking and admin on purpose, because a change usually moves one block and not another, and a single "how many hours" hides which. The fourth question is not decoration: what teachers would do with two hours a week tells you whether the saving is worth having, and it is usually the sentence that ends up in the staff meeting. The fifth is the honesty check, and a "no" with a reason is the most useful answer the survey can produce.
Before the change,
and a full term after it.
The first survey is taken before anything changes — before the platform, before the written curriculum, before the assessment redesign — and its purpose is the baseline. Taking it a month into the change is the commonest mistake, because the first month is the worst: everyone is slower with a new tool, and the baseline you record is the trough rather than the old normal. The second survey is a full term later, in the same week of the term, because the same week of a different term is the only fair comparison; the week before exams and the week after half-term are not the same week.
Two surveys are enough for a decision. A third, a year on, tells you whether the saving held or whether the old habits returned, which they do in schools that changed the tool and not the way of working. If you run only two, run them a term apart. If you run three, keep the same week each time.
Hours freed,
into capacity or evenings.
Add each teacher's three numbers before and after, and take the difference: that is hours freed per teacher per week. Multiply by the number of teachers and by the weeks in the term for the school's total. Then decide, as a school, what the hours are for, because the same number means different things. As an illustration with placeholder figures: a teacher on [18] contact hours reporting [12] hours a week of preparation, marking and admin before and [7] after has [5] hours freed. Turned into teaching, that is [5] more contact hours from the same salary, roughly a [28]% increase in that teacher's capacity, which for a school with a waiting list is a class it can now run. Turned into evenings, it is a teacher who stays.
Say which it is. A school that quietly converts freed hours into more classes has increased workload back to where it was and will get the honesty-check answer it deserves next term. A school that says "three of the five hours are yours, two are a new class, and here is the pay for it" has made a decision teachers can see. The number is only useful if it leads to one.
Six answers are six answers,
not a percentage.
A school of six teachers will get six pairs of numbers, and the honest way to report them is as six pairs, not as an average dressed up as a finding. One teacher whose hours doubled because they are teaching a new level tells you something the average hides. A teacher whose hours barely moved because they never used generation tells you the saving is a habit, not a licence. Look at the spread, read the fourth and fifth answers alongside the numbers, and resist turning six responses into a headline. If a school wants a number to put in a sentence, the sentence should carry the context: how many teachers, one term, self-reported, and what changed.
That is the standard we hold ourselves to on this site, and it is the reason the previous article's claim is one sentence with four caveats rather than a figure on the home page. If your school runs this survey, before and a term after, we would like to see the result — whatever it shows — and we will publish what schools find whether or not it matches ours.