Your LMS Cannot Tell You Which Student Is Drifting

Every school has the meeting. Mid-March, a spreadsheet on the screen, a column of predicted grades, and someone asking why 11B is two-thirds of a grade behind 11A. Nobody in the room can answer, because the data says what and never which topic. So the action point becomes “more revision”, and everybody writes it down.

The gap is not effort and it is usually not teaching. It is that the school’s systems record marks rather than mistakes. This is written for principals, heads of secondary and heads of department: what your LMS gradebook can and cannot tell you, what topic-level tracking changes, and how to build a data habit that produces named actions instead of adjectives.

What a gradebook actually knows

An LMS gradebook holds numbers a teacher entered after marking. That means three things, and every school data meeting runs into all three:

• It only knows what was marked. The practice a teacher set but had no time to mark simply does not exist in the data, and that is usually the majority of practice.
• It aggregates away the answer. A 62% on a Physics paper is six correct questions and four wrong ones, and only the four matter. The gradebook stores the 62.
• It is one teacher’s judgement per column. Comparing 11A and 11B through it compares two marking styles as much as two cohorts.

None of this is a criticism of the LMS. Distributing work and recording grades is its job, and it does it well. It was never built to hold a syllabus. The layer that is missing is a different layer, and no amount of LMS features closes it.

Topic-level tracking, and what it changes in practice

The shift is small to describe and large to live with. Instead of storing a mark per assessment, the system stores a result per question, and every question carries its syllabus point. Everything useful falls out of that one decision.

The four screens leadership actually needs

Most school analytics fail by showing activity: logins, submissions, minutes. Activity data answers “is the software being used”, which is a procurement question, not a leadership one. These four answer leadership questions.

1. Who needs an adult this week — and why
   The reason is the feature. “Accuracy down 14 points over three weeks, all of it in one topic” produces a conversation. A bare list of names produces a scroll.
2. The department view, weakest subject first
   Each subject’s average, how much of the syllabus has been touched, and the exact topics to re-teach. This is the agenda for the departmental meeting, written by the data rather than by whoever prepared the slides.
3. The trend over eight weeks, not the snapshot
   One mock is a photograph. Eight weeks of accuracy is a direction, and direction is what tells you whether last month’s decision worked.
4. The teacher view, honestly
   Classes, students, what has been set, class average, and who has set nothing in a fortnight. Uncomfortable, and far kinder than discovering it in August.

Why the timing matters more than the depth

Schools tend to buy analytics for richness and then use them once a term. The value is almost entirely in the opposite: shallow data every week beats deep data after the mock, because only one of them arrives while there is still a term left to act.

The 2026 exam year made this concrete across the region. When May/June exams were cancelled in the UAE, Kuwait, Bahrain and Qatar and grades came from a Portfolio of Evidence, the schools that coped had a continuous, per-student record of marked work already sitting there. The ones that did not spent half-term building one from filing cabinets. That was a stress test of exactly this capability, and it sorted schools into three very visible groups.

Building the habit without buying anything new

Three of these cost nothing and are worth doing whatever you decide about software.

1. Tag your existing assessments to syllabus points
   One department, one term. Even on paper, recording which syllabus point each question tested turns your existing mocks into topic data. It is tedious once and useful forever.
2. Change the standing item in the data meeting
   Replace “review grades” with “name the three weakest topics per subject and who is re-teaching them, and when”. The meeting gets shorter and the minutes get actionable.
3. Give parents topics, not adjectives
   Two named strengths and two named weaknesses per subject at every parents’ evening. Parents can act on a topic; nobody can act on “could work harder”.
4. Then automate the part that is unsustainable
   The reason schools abandon topic tracking is marking volume, not principle. That is the point at which software genuinely helps — and the marking arithmetic is worth reading first.

For what it is worth, this is what PapaMarks exists to produce: every one of the 224,000+ past-paper questions is tagged to its syllabus topic, answers are marked as students hand in, and the school office turns that into the four screens above — the attention list with reasons, the department view weakest-first, the eight-week trend, and a per-student page with the six weakest topics on it. Leadership visibility is part of a school licence rather than a premium tier, because a school that cannot see the data will not build the habit.

Frequently asked questions

The schools that move a grade are rarely the ones that meet about data most often. They are the ones whose data names a topic, a class and a week — and whose meetings end with somebody re-teaching something on Tuesday.

Last reviewed: August 2026. Corrections welcome.