How Coaching Institutes Can Spot At-Risk Students Early, Before the Drop Year

You spot at-risk students early by watching a few signals every week, not just the monthly mock score: the attendance trend, test scores falling below the student's own earlier level, missed tests, and recorded lectures left unwatched. When two of these slip together, a teacher should call, weeks before a full-length test confirms the problem.

This guide is for owners and academic heads of NEET, JEE and other competitive-exam institutes. It covers why slipping students are found so late, which signals are worth watching, a weekly review routine that fits into twenty minutes, and how The Academix early-warning tags and alerts support that routine without pretending to read anyone's future.

Why do coaching institutes find out too late?

In a batch of 150 or 200 students, no single person sees the whole student. The physics teacher sees physics scores, the chemistry teacher sees chemistry, and the front desk sees attendance registers nobody reads. Full-length mock tests come once or twice a month, so the first time anyone sees the whole picture is a bad all-subject score, and by then the slide is a month old.

Students also hide it. A repeater in a drop year is under pressure from home and from themselves, and the usual response to a bad patch is to go quiet: skip the weekly test, sit at the back, stop asking doubts. Adult aspirants preparing for UPSC or CA rarely tell anyone they are falling behind.

A drop year is short. Between one result and the next attempt there are only a few months of real teaching, and a student who loses six weeks to a slump in October rarely gets them back. Catching it in week two instead of week eight is the whole game.

Which signals show that a student is slipping?

Attendance trend, not the total

A student with 85% attendance for the term can still have attended half the classes in the last three weeks. Look at the recent trend. A sudden drop after steady attendance says more than a low but stable figure.

Scores falling against the student's own level

Compare each student with their own recent average, not with the topper. A student moving from 60% to 45% across three weekly tests is in more trouble than one who has always scored 45%, even though the second looks worse on the rank list.

Missed tests

This is often the earliest signal. Students skip tests they expect to do badly in, so two missed weekly tests in a row usually come before the bad score, not after it. Count tests attempted against tests released.

Recorded lectures left unwatched

For hybrid and recorded batches, lecture completion shows whether a student is keeping up with the syllabus between tests. A student who has not opened the last ten lectures is behind, whatever their last score says.

Recent activity

When did the student last sign in, submit an assignment or take a practice quiz? A week of silence from a student who used to be active is worth a question.

One signal on its own is often noise: a student was ill, a family wedding, a bad paper on a hard chapter. Two or more together, over two or more weeks, is a pattern.

How should a weekly at-risk review work?

The review is what turns data into help. It does not need a meeting room or a new role. Use this checklist:

• Same day every week, twenty minutes, with the batch coordinator and the subject teachers for that batch
• Start from the list of students showing two or more signals, not from the whole batch
• Give each flagged student one owner, usually the teacher who knows them best
• Agree one action: a one-to-one conversation, a doubt session, a make-up test, or a call to the parent
• Write down what was done and the date to check again
• Next week, close the case if the student is back on track, or escalate to the academic head if not
• Note false alarms such as illness or travel, so the same student is not chased twice for the same reason

What should a teacher say to a student who is slipping?

Start with what you saw, not with a judgement: you missed the last two Sunday tests and your chemistry scores have dropped, what is going on? Then listen. The reason is often practical: a chapter they never understood, a clash with school exams, a health problem, trouble at home. Agree on one small step with a date, such as attempting next Sunday's test even if they feel unprepared, or coming to a doubt session on Wednesday. A student who is told the plan will try it; one who helped make it usually does.

When should parents be told?

It depends on the student. For school-age students and most NEET and JEE aspirants living at home, parents are part of the solution, and a parent who hears early is far less angry than one who hears at result time. For adult aspirants, a call home can feel like a breach of trust. A good rule is to talk to the student first, then decide together whether the parent should know. Whatever you decide, decide it as a policy, so every student in a batch is treated the same way.

How do early-warning tags and alerts work in The Academix?

The Academix tags each student On Track, Needs Attention or High Risk from five things it already records. Over the last 90 days, test scores count for 30%, attendance 25%, tests attempted out of tests released 25%, lecture completion 10% and recent activity 10%. A total of 70 or more is On Track, 40 to 69 is Needs Attention and below 40 is High Risk. The class student list in the institute panel shows the tag with its reasons, for example "Attendance 62%" or "Missed 3 of 5 tests". Tags appear without any setup.

Each test result also carries a Performance Summary built from the student's last five tests of that type: an overall level, what changed compared with the class average, the weakest topics and a "What to do next" list. Students see the same summary in their own portal, so the conversation with a teacher starts from the same page.

On top of the tags, an owner can switch on four email alerts in Settings, under Alerts: low attendance, score decline, test not attempted and overdue intervention. Each has its own limit. Every day at 7 AM IST the platform checks every student and emails the teachers, staff or owner you chose, after yesterday's attendance has been finalised. Each alert has a switch to copy the student's parents linked in the parent portal. An alert opens an intervention, staff resolve it once someone has followed up, and if it stays open past your limit the overdue alert fires.

What it does not do: it is not AI and makes no prediction about who will leave; the tag cut-offs are the same for every institute; alert settings apply to the whole institute rather than class by class; alerts go by email only, with no SMS, WhatsApp or push; and staff cannot open an intervention by hand. Early-warning alerts are on every plan. Plans start from ₹1,500/month; the final price is quoted to your requirements.

In practice the 7 AM email becomes the agenda for the weekly review: the list of students to discuss is already there, with the reason next to each name, and the open interventions list shows what nobody has followed up yet.

Frequently asked questions

Does The Academix predict which students will drop out?

No. It tags students from fixed rules over test scores, attendance, test participation, lecture completion and recent activity, and shows the reasons. A teacher decides what the tag means for that student and what to do.

Is the risk tag worked out by AI?

No. It is a weighted score with fixed cut-offs: 70 or more is On Track, 40 to 69 is Needs Attention, and below 40 is High Risk. The same rules apply to every student, so a teacher can always see why a tag was given.

Can we change the cut-offs or the alert limits?

The tag cut-offs are the same for every institute. The four alerts are yours to configure: each has its own on/off switch and its own limit, such as an attendance percentage or a number of days, set for the whole institute.

Do alerts go out on WhatsApp or SMS?

No, alerts are sent by email only, every day at 7 AM IST, to the teachers, staff or owner you choose. A student triggers each alert at most once every seven days.

What happens with a new student who has no test data yet?

A student with too little data shows No Data instead of a guessed tag. If one of the five signals has no data for a student, it is left out and the others are scaled up.

If you want to see the risk tags, a Performance Summary and the alert settings on a demo institute, book a demo and we will talk through which limits suit your batches.