A student arrives at 8:40 on a Monday. Their first class starts at 9:00, they enrolled online eleven days ago from another country, and nobody in the building has heard them speak. Placement testing in a language school almost always happens under that kind of pressure — twenty minutes, a room that doubles as the staff kitchen, and a Director of Studies who still has a timetable to publish.
Get it right and the student settles and stays. Get it wrong and you find out in week two, when a teacher says someone is lost, or visibly bored, and a group has to be rebuilt around one person.
This post covers the whole lifecycle, not just the test: placing a new student before their first lesson, what a placement decision must record so it can be defended later, and the part nobody writes about — what breaks when a student changes level mid-course.
The twenty minutes before the first class
Language academies run rolling intakes. There is no single September morning when everybody arrives and everybody is tested. There is a Monday, every Monday, with anywhere from two to thirty new students from different countries, a good share of them booked through agents who have already told them which level they will be in.
So placement happens in the gap between reception and the first lesson. That constraint shapes everything: the test has to be short, markable on the spot, and decided by someone who can be interrupted three times while deciding.
Two things go wrong here, and they are opposites. Either the assessment is too thin — a written exercise, no speaking — and a student who reads well but cannot hold a conversation lands above where they belong. Or it is too heavy for 8:40, so it gets skipped for students who "obviously" belong somewhere, and the obvious turns out to be the agent's opinion.
What a placement decision actually decides
On the form it is one field. In the operation it is not. The level decides the group, and the group decides the timetable, the classroom, the teacher, the classmates, the coursebook and the point in the syllabus the student walks into. In most academies it also decides a line on the invoice, because general and intensive are rarely priced the same.
So a placement is worth storing as a decision with evidence attached, not a value typed into a cell. When a student disputes their level in week two, when an agent asks why their client is a level below what was promised, or when an inspection asks how levels are assigned, what helps is a dated record of what was assessed, by whom, and what was concluded.
Who makes the call, and on what evidence
Placement is an academic decision made under commercial pressure, so naming the decision-maker in writing is the cheapest safeguard you have: the Director of Studies or an academic coordinator, sometimes a senior teacher on a rota, and never reception staff working from the booking form alone.
The evidence worth recording
A defensible placement rests on more than one signal:
A written or online component. Fast to mark, easy to standardise, and on its own not enough.
A short spoken interaction. Five minutes with a teacher separates two students whose written work is indistinguishable. It is the component that gets dropped when the intake is large, and the one that prevents the week-two problem.
Prior evidence. A certificate, a report from a previous school, the level the student declares. Useful context, but a claim until you test it — and an agent's claim is a commercial statement made months earlier.
The decision itself. The level assigned, the date, the person who assigned it, one line of reasoning. That last line takes ten seconds and it is the only thing that helps you six weeks later.
The pressure to place into whatever group has space
This is the real corruption risk in placement. A student tests between two levels; one group has four free seats and the other is full. The commercial answer and the academic answer point in different directions, and the person under pressure is usually the one who also has to fill the timetable.
You cannot design that pressure away, but you can make it visible. Record the placement decision separately from the group allocation, so "assessed at this level" and "placed here because it was the only group running" stay two facts rather than one blurred one. When borderline students are consistently absorbed by whichever group has room, you have evidence for opening another group instead of an argument based on impressions.
CEFR levels as an operational structure, not a label
Most academies describe their groups using the Common European Framework of Reference. As a marketing label it means little. As an operational structure it does real work: a shared vocabulary between teachers, a way to sequence groups, a basis for comparing progress across two campuses.
The requirement is that a level means the same thing on Tuesday in Cork as on Thursday in Dublin — descriptors written down, sub-levels defined rather than improvised, and the mapping between your internal group names and the framework living somewhere other than in the Director of Studies' head. If you need to state what your levels represent formally, on certificates or for accreditation, confirm the requirements with your national accrediting body rather than copying another school's wording.
Re-levelling mid-course: the part nobody writes about
Here is the assumption that breaks most software sold to language academies: that a student enrols in a group and stays there until the course ends. That is how a school year works, not how a language academy works.
A student who arrives in January can move up in week three, move again in week nine, extend twice, and leave in July having been in four groups with three teachers. That is a normal, successful student — and the hardest thing to represent in a system designed around semesters, which is why academies end up keeping the truth in a spreadsheet next to the software they pay for.
What a level change fragments
Move one student on a Wednesday and you have split several records at once:
Attendance. Their history now sits partly against the old group and partly against the new one. Any report that assumes one student equals one group will double-count or lose the first three weeks.
The timetable. The new group may run at a different time, in a different room, possibly at a different centre. Two teachers now hold partial knowledge of the same student.
The group they left. Numbers drop below the point where it is viable, or the group they joined tips over its cap, and one decision forces a second.
Assessment history. Marks earned against one level's criteria sit alongside marks from another, and combining them means nothing unless the level is attached to each result.
The invoice. If the new group carries a different price or end date, finance needs to know today, not at month end.
None of these is difficult alone. All of them at once, for a handful of students a week, is where the administrative hours go.
Running the change without losing the history
A level change is a new period in a continuous record, not a rewrite of the old one. In practice:
Record the trigger — a recommendation, a re-test, a progress review, a student request — with who raised it and when.
Confirm it academically. Whoever owns placement owns re-levelling. A teacher recommends; the academic lead decides.
Set an effective date instead of editing the level in place. Everything before it belongs to the old level, everything after to the new one.
Close the old enrolment and open the new one, keeping both visible on the profile. Never delete the first group — that is the history someone will ask about.
Tell everyone the same thing: the student, both teachers, reception, and finance if money changes. This is where it fails, when a change agreed on Wednesday reaches finance in October.
Check the group they left. A move out is a capacity event for two groups, not one.
Re-testing on a schedule, not on complaint
If the only trigger for reassessment is somebody complaining, you catch only the extremes. The quiet student sitting a level below their ability finishes the course, learns less than they could have, and does not recommend you.
A light, scheduled review — at fixed points in the course, or at each booking renewal — turns re-levelling from an exception into a routine. It need not be a full sit-down test; a teacher's structured judgement against written level descriptors is usually enough to flag who deserves a closer look. What matters is that it happens on a calendar and that the outcome, "no change" included, is written down.
What a student's record should look like after two level changes
Open the profile of a student who has moved twice. Without asking anybody, you should see:
The original placement: date, level, who decided, what evidence.
Each level change: date, direction, trigger, who approved it.
Attendance across all three groups, both as one continuous figure and split by group.
Assessment results, each tagged with the level it was earned at.
The financial consequence of each move, where there was one.
Which teachers taught them, and for which weeks.
If producing that means opening three spreadsheets and messaging two teachers, the record does not exist — it is reconstructed from memory every time somebody asks.
What this looks like with a system built for rolling intakes
The difference is not a better placement test. It is that the level, the group and the enrolment are separate dated things that can change independently while the student's history stays in one place.
KMPUS holds classes, levels, academic periods, attendance and evaluations inside one student profile, so a level change is a new period on an existing record, not a new row in a new sheet. Because the platform was built for language academies with rolling intakes and continuous level changes rather than adapted from a school-year product, moving a student mid-course is an ordinary operation, not a workaround.
The lifecycle starts earlier, too. Online enrolment and document collection means the student's file, their declared level and their agent's booking exist before they walk in on Monday, so the twenty minutes at reception go on assessing rather than typing. And when a move changes what a student owes, payments and family statements live in the same system as the academic record — the join spreadsheets never survive.
Across more than one site — NED College in Dublin and Limerick, Erin School of English in Cork and Dublin — level consistency and clean histories stop being a preference. Their customer stories are worth reading if a second campus is on your horizon.
Frequently asked questions
How long should a placement test take?
Short enough to run before the first class of the day: a written or online component plus a brief spoken interaction. A process that cannot fit the gap you have on a Monday gets skipped on the busy weeks, which are the weeks it matters most.
Does a placement test need a speaking component?
For a language academy, yes. Written tasks are easy to standardise, but they misplace students whose productive and receptive skills are far apart, and a few minutes of conversation resolves most borderline cases.
What happens to attendance when a student changes level?
It should split by group and still total correctly for the student: weeks one to three against the group they left, week four onwards against the group they joined, and any whole-enrolment report combining both without counting either twice.
If placement and re-levelling are held together by a shared spreadsheet and a good memory, see what the same workflow looks like in a system built for it — start a free trial of KMPUS, no credit card required.
