Almost every coaching institute in India is running the same stack, and almost nobody chose it. It assembled itself: Zoom or Meet for classes, Google Forms for tests, a WhatsApp group per batch, an Excel sheet for fees, and a physical register for attendance. It cost nothing, it worked under real pressure, and it deserves more respect than it usually gets from people trying to sell you something.
It also has a specific failure point. The failure is not the number of students — plenty of large institutes run on it badly and plenty of small ones run on it beautifully. The failure is the number of questions you have quietly stopped asking because answering them costs a morning.
The stack, honestly: what each free tool is genuinely good at
Start by being fair to it, because a comparison that pretends these tools are bad is a comparison you should not trust.
- Zoom and Meet deliver reliable video to a hundred people on unreliable connections. That is world-class engineering, and you are getting most of it free.
- Google Forms with Sheets is the fastest way ever built to make an MCQ test, send it, and see scores. No training, no cost, no setup.
- WhatsApp is where your parents already are. There is no app to install and no adoption problem, and its read rates are something no software vendor can promise you.
- Excel is infinitely flexible, and every institute has one person who is genuinely good at it.
- The paper register is fast, needs no training, and works when the power is out.
Nothing on that list is weak. The problem is not quality. It is that none of these tools knows the others exist, and the institute's most valuable questions live precisely in the gaps between them.
The joins nobody does: attendance, marks and fees never meet
Almost every decision worth making in an institute requires two of these systems to be read together. Consider what you would actually like to know on a Monday morning.
- Which students in this batch have attendance below seventy percent and also scored under the batch average in the last two tests?
- Of the students who left after the second installment, how many had already stopped attending regularly before they stopped paying?
- Which faculty's batches show the widest gap between attendance and test performance — the classes students turn up to and still do badly in?
- Which parents have both an overdue installment and a child whose attendance is sliding? Those are the calls to make today, and they are different calls from the ones to families with only one of the two problems.
Every one of those is ten to twenty minutes of manual cross-referencing per batch, done by matching names between a register, a spreadsheet and a folder of test results. Which means nobody does it. Which means the information does not exist inside your institute even though all of the underlying data does.
The free stack does not lose your data. It loses the relationships between your data, which is where every useful answer lives.
Present on a video call is not attendance
The participant list on a video call feels like attendance and is not, for four reasons that all bite at once.
- Joining is not attending. A student who connects, mutes and walks away appears on the list exactly like the one who sat through the whole class.
- Duration exists but nobody uses it. It is available in a report somebody must download per meeting, per day, and then merge by hand.
- Names do not match your roster. Papa's phone, a nickname, an initial, an iPhone. Reconciling that to a class list is a person's evening.
- Nothing flows anywhere. Even a perfectly reconciled list ends up in a folder, not on the student's record and not in front of a parent.
For attendance to be worth having it needs four properties: computed against the classes actually held rather than a nominal total, attached to the student, visible to the parent, and produced without anyone typing. Live classes that run inside the same system as the roster deliver that by construction — who joined and for how long is captured automatically and lands in the attendance record itself, so a percentage means something and an absence can reach a parent the same hour instead of at the end of the month.
For offline batches the equivalent move is a biometric device. The gain is not really the saved minute; it is the end of the argument. A parent told that his child missed the fourteenth is a completely different conversation when the record was made by a fingerprint reader at the door rather than by a teacher's recollection.
Tests: instant MCQ grading versus a folder of scanned answer sheets
Credit where it is due: for a single MCQ test, Google Forms is excellent and instant. If that is genuinely all you need, you do not have a testing problem.
It stops in four places. It holds a test, not a question bank you can tag by chapter and reassemble into next month's paper. It has no answer for written papers, and serious exam preparation in India is not pure MCQ — subjective answers become a pile of scans and a teacher's Sunday. It cannot produce a rank or a percentile without somebody building formulas in a spreadsheet, per test, forever, and rank is the number your students and their parents actually care about. And the score ends its life in a sheet, unattached to the student, invisible to the parent, incomparable across tests.
Testing inside the platform changes the shape of that work rather than eliminating it. MCQs grade themselves. Written answers get auto-scored against your own model answer, with the teacher keeping the final say — the value is that a teacher reviews and adjusts rather than starting from a blank page at eleven at night. Ranks and percentiles publish the same day. And the result lands on the student's record, next to attendance and fees, where it can finally be read alongside them.
Where parents actually look, and what they see today
Be honest about what a parent experiences right now. A group with a hundred and eighty members, mostly forwards and holiday notices. An occasional photograph of a marks list. A fee reminder that arrives without context.
What parents want is boring and unchanging: is my child actually attending, how did the last test go, what do I owe and when, and what is the timetable. Today every one of those questions becomes a phone call to your front desk, which is a large part of why your front desk is always busy in the first week of the month.
A portal that a parent signs into with a one-tap SMS OTP or an email link — showing their own child's fees, attendance, results and timetable, and nothing about anybody else's — changes two things. It removes a great many calls. And it shifts the emotional register from asking to checking, which matters more than it sounds when the next message you send is about money. The fee conversation is simply easier when the amount due is on the same screen as the attendance and the results it paid for.
None of this means killing the WhatsApp group. It is genuinely good at announcements and will remain so. It just should not be your system of record.
What breaks first as you add batches
Institutes do not outgrow the free stack all at once. They outgrow it in a fairly predictable order, and the driver is batches and staff rather than any magic student number.
- Attendance goes first. One person can hold four batches in a register. Nobody holds twelve, and the percentages stop being trustworthy long before anyone admits it.
- Fees go second. The spreadsheet and the receipt book stop agreeing, and no one can say which is right.
- Tests go third, usually the week you need a rank across three parallel batches sitting the same paper.
- Parent communication goes fourth, and it goes badly — the day a message meant for one batch reaches another is the day the stack has publicly failed.
- The owner goes last. You stop knowing things directly and start asking three people, and every answer takes a day to arrive.
The clearest signal is not any of those individually. It is noticing that you now make decisions on feel — which batch to expand, which teacher to keep, which fee plan to change — because assembling the evidence would cost more than the decision seems to be worth.
The switching cost, stated plainly
Moving off the free stack is not free, and anyone who tells you otherwise is selling. The real costs are these.
- Data entry. Students, guardian phone numbers, batches, fee schedules. Tedious. Restrict it to current students; historical records can stay where they are or come later.
- Staff habit. Expect two to four weeks of people doing it the old way alongside the new way. Budget the patience rather than being surprised by it.
- The parent-facing switch. Announce it once, clearly, then hold the line. If you keep answering portal questions on WhatsApp for a month, you have two systems and the benefit of neither.
- Timing. Do it between sessions or at a batch start. Doing it in week three of admissions is how a good decision gets blamed for a bad month.
What comes back is mostly hours — attendance not typed, papers not tabulated, receipts not reconciled, the same four questions not answered on the phone twenty times a week — plus a set of answers you currently do not have at any price. Whether that trade is worth it depends entirely on your size, which is the honest last question.
Who should stay on the free stack
If you run one or two batches, teach most of the classes yourself, know every student's parent by name and collect fees in two lumps, stay exactly where you are. You are the join. Every relationship the software would maintain, you already hold in your head more accurately than any system would, and adding process at that size buys you nothing but process.
The same applies if your entire model is a single online cohort with no offline component and no written papers. Video plus forms plus a group is genuinely a coherent stack for that shape of business.
Move when more than one person handles money, when there are more batches than you can hold in your head, when parents ask questions you cannot answer while still on the phone, when written tests need ranks across parallel batches, or when you catch yourself deciding on instinct because the evidence is too expensive to gather.
The trigger was never a student count. It is the moment you stop being the join between your own systems — and the only real question after that is how many months you spend being the join manually before you admit it.