This cluster is about delivering online, not about deciding to. Whether a school should open online, on premises or both is a business question and it is answered in the starting cluster. What follows assumes the decision is made and asks the harder question: given that the class is happening on a screen, what has to be different about it.
A great deal, as it turns out, and almost none of it is technical. Schools that moved online and were disappointed usually changed the venue and nothing else — same lesson plan, same teacher-fronted rhythm, same assumption that being present means participating. The result is a class where four people have their cameras off and one person is doing all the talking, and where attendance decays in a way everybody attributes to motivation. It is not motivation. It is design.
The purpose of a language class is that learners produce language. Everything else — explanation, correction, grammar, materials — exists to make production possible. In a room, a class that is eighty per cent teacher still gives students the corridor, the pair next to them, the mumbled exchange before the bell. On a screen there is no corridor. If the session is eighty per cent teacher, the students produced almost nothing, and they will feel it as boredom without being able to name why.
So the first metric for an online language class is not attendance or satisfaction. It is the proportion of the hour in which students are speaking, and the honest way to find out is to time it once. Most teachers are startled. Restructuring to fix it means fewer explanations and more tasks, pairs and small groups — which is what breakout rooms exist for and why a platform without them is a poor fit for language teaching specifically, whatever else it does well.
On SprintUp a teacher applies a saved group set to a live class, and each group gets its own breakout room. The groups are copied when the activity starts, so whatever happens during the call never silently rewrites the saved set, and students move themselves into their room rather than waiting to be dragged. Only one group activity can be active in a call at a time, which is enforced rather than left to discipline. The first article covers how to actually use them, including the timing that makes them work and the reason unstructured breakouts fail.
Language learning has a long stretch where progress stops being visible, and most people who quit, quit there. In a building, the habit and the other people carry a proportion of them through it. Online, leaving costs nothing and happens silently: no conversation, no cancellation, just a student who stops appearing and a school that notices a month later.
Which means an online school has to manufacture, deliberately, what a building supplies free. Stable groups rather than rosters that reshuffle, so students know who they are learning with. A teacher who contacts an absent student the same day rather than at the end of the month — the single highest-return habit in online teaching, and the subject of the absence follow-up article. Visible progress that is not only a score. And a course with an end date, because an open-ended subscription invites a monthly decision to continue.
The platform can help with noticing. In a live class SprintUp knows how long each student was actually connected and uses it to suggest a register entry against a policy the school sets — by default, connected for at least 75% of the call counts as present, and joining more than ten minutes late counts as late. The suggestion only pre-fills. A teacher confirms or overrides it, because a dropped connection is not the same as an absence and a register that got that wrong would be wrong about something families care about.
Most schools offering something online did not decide to. A student asked to join remotely, it was easy to say yes, and a year later there are classes where some people are in the room and some are on a screen — the hardest format to teach well and the one nobody chose. The second article is about doing it on purpose instead: which elements genuinely belong in a live session and which are better asynchronous, how to assess across both without doubling the work, and how to keep a single group coherent when its members experience the course differently.
The short version, because it is worth having before you read on: synchronous time is expensive and should be spent on the things that need other people — speaking, negotiation of meaning, feedback in the moment. Anything that a learner can do alone at their own pace, including most input and most drilling, is a poor use of a live hour and a good use of the week between them.
An online language school needs a way to hold live classes, a place for materials and coursework, a record of who is enrolled and who attended, and a way to take money. That is the list. Schools routinely buy far more than that — separate tools for scheduling, homework, messaging, assessment and marketing — and then spend the year moving information between them by hand, which is a cost nobody counted at purchase.
The third article works through the stack and the question that decides most of the cost: whether these things are one system or five connected by integrations that somebody has to maintain. It is the same argument as the red flags article makes about "there is an integration for that", applied to the specific shape of a language school.
Run a live class that produces speech, design a hybrid course rather than drifting into one, and buy only the four things you need. Each article ends with a checklist.