Blended by design,
not some people on a screen.
The word covers at least three arrangements, and only one of them is the subject of this article. There is separate cohorts, where the same course runs as an in-room class and as an online class, each taught as itself. There is a physical school with online reach, serving students who could never attend in person. And there is blended design, where a single course deliberately places some elements live and some in the week between, for everyone.
This article is about the third. The choice between the arrangements, and the fourth version that arrives by accident — one class, some people in the room, some on a screen, decided lesson by lesson — is covered in online-first or premises. The short warning worth repeating: the accidental version is the hardest to teach and the one nobody chose, and if you recognise your school in it, choosing one of the three is the single biggest improvement available.
What deserves
a live hour.
One test settles most of it: does this activity require other people to be present at the same moment? If yes, it belongs in the live session. If a learner could do it alone, at their own pace, at eleven at night, then spending scarce synchronous minutes on it is a waste of the most expensive thing the course has.
The second row is the one schools resist, because moving explanation out of the live hour feels like abandoning teaching. It is the opposite. A recorded or written explanation can be paused, repeated and revisited by the student who needs it three times, while the one delivered live is heard once at a speed that suits roughly one person in the room. What the live hour then gains is the thing a recording can never provide, which is somebody to talk to.
The last row has a practical payoff worth naming: moving written assessment out of live time recovers hours across a term that most schools did not realise they were spending on invigilation.
One course,
experienced differently.
The risk in a blended design is that the course fragments: the diligent do the asynchronous work and arrive ready, the rest turn up having done nothing, and the live hour splits into two classes happening in the same room. Left alone this widens every week until the teacher is either re-teaching the input — which wastes the design — or leaving half the group behind.
Three things hold it together. First, make the asynchronous work small and genuinely necessary rather than large and optional; a ten-minute task that the live session visibly depends on gets done, a forty-minute one that could be skipped does not. Second, begin the live hour with something that only works if the preparation happened — a task that uses it, not a quiz that tests it, because the first is a reason and the second is a punishment. Third, make the state visible: know before the session who has done what, so that you walk in informed rather than discovering it in the first five minutes.
That third point is where a platform earns its place, because the alternative is a teacher chasing by message. Coursework and attendance sitting against the same class record means the preparation is a fact you can look at rather than a hope, and the reporting article covers which of those numbers is actually worth watching weekly.
Speech live,
writing in the week.
Assessment in a blended course should follow the same allocation as everything else, and it usually does not. Speaking has to be assessed live, because that is what it is; there is no asynchronous substitute that measures the same thing. Writing should be assessed asynchronously, marked properly, and returned before the next session while the student still remembers producing it — feedback that arrives a fortnight later is a grade, not feedback.
The workload objection is real, and it is the reason most schools assess less than they should. Marking written work is the largest recurring cost in a language teacher's week. SprintUp can assist with it — written answers can be marked with AI support, with the teacher reviewing rather than the machine deciding — and the honest framing matters here: it is a first pass that a teacher corrects and signs off, not automatic grading, and a school that presents it to families as anything else is storing up a problem. Used that way it returns time to the part that cannot be delegated, which is deciding what to do about what the marking revealed.
Hybrid is more preparation,
not less.
A blended course has to be built twice over: the live sequence and the materials that carry the week between. Schools adopt it expecting efficiency — fewer contact hours, therefore cheaper — and discover that the contact hours fell while the preparation rose, which is a bad trade if nobody planned for it and a good one if they did, because the materials are reusable and the contact hours are not.
So budget the build explicitly, count it as a one-off investment against several terms, and resist redesigning it every cycle. The mistake that makes hybrid genuinely unaffordable is treating the asynchronous half as something to produce fresh each week. Build it once for a level, improve it in the places the evidence says are weak, and reuse the rest. The workload article covers how to count preparation honestly, which is the only way to know whether the trade went the way you intended.