One campus, nine spreadsheets: when a school actually needs an ERP

A four-hundred-student school does not run on software. It runs on a clerk who has been there eleven years, a vice-principal who remembers everything, and a folder of spreadsheets that only make sense to the person who built them. It works. That is the awkward part of this conversation — it genuinely works, right up until the person who holds it all together takes leave in March.

Meanwhile the ERP being quoted to that school was designed for a group with fourteen campuses, a central admissions office, hostels, and a transport fleet with GPS on every bus. The demo is impressive and almost none of it is about the school in the room. So the principal signs nothing, goes back to the spreadsheets, and is quietly told once a year that they are behind the times.

The honest question is narrower than the sales pitch: which specific jobs at a single campus are being done badly by spreadsheets, and which ones are being done perfectly well.

The nine spreadsheets, and who owns each one

Every single-campus school converges on roughly the same set. Count yours.

• The student master — office clerk, and the only complete list of who is enrolled
• The fee collection register — accounts, updated daily from the receipt book
• The defaulters list — accounts, rebuilt by hand every month from the register
• Attendance consolidation for the annual return — the vice-principal, in March, from twenty class registers
• Marks entry sheets, one per subject — subject teachers, emailed to the exam in-charge
• The report card sheet with the formulas — one teacher who built it years ago and is now structurally indispensable
• Bus routes and stops — the transport in-charge, on his own laptop
• Staff attendance and leave — admin, entirely separate from everything else
• The parent contact list — the front desk, currently circulating in three versions

The problem is not the number of files. It is that each one is a private copy with its own version of the truth, and reconciliation is performed by a human being with a good memory. A child who left in August is still in the transport sheet in November. A phone number updated at the front desk never reaches the class teacher. The exam in-charge is working from a class list that predates two admissions.

And every one of these files has a successor named something like final_v4_updated, sitting on a laptop that will leave the campus one day with the person carrying it.

What Tally is doing for you, and what it is not

Be fair to Tally, because most schools use it and most schools should keep using it. It is the accounting standard in this country for good reasons: your auditor knows it cold, your statutory books come out of it, your GST position is built in it, and it does not lose money. Nothing written here is an argument for replacing it.

But Tally understands ledgers, not students. It has no concept of a batch, a fee structure made of components, a per-student instalment plan, a staff-ward concession, a late-fee rule, an attendance session or a marksheet. Schools that try to push the whole fee operation into it end up creating a ledger per student, which becomes unmanageable somewhere in the low hundreds and still cannot answer the questions that matter: why this fee is unpaid, when someone last spoke to the family, and which class is worst affected.

The sane division of labour is to let each tool do what it is good at. The fee lifecycle — structures, bills raised, collection, discounts, receipts, outstanding — belongs in the system that knows who a student is. The books belong in Tally. Ask your accountant what summary they want posted and at what frequency, and hand them that. What you should not do is maintain the same information twice by hand, which is what a school is doing every time the accounts clerk retypes the day's receipts into a second system.

Where a single campus genuinely loses time

A single-campus school is not drowning every day. The load is spiky, and it lands hard on two or three people who cannot be replaced during the spike.

• The first six weeks of the session — admissions, fee structures for the year, the first big collection, and new students missing from half the lists
• Every fee cycle — reconstructing the defaulters list by hand, then chasing it with a phone in one hand and the register in the other
• Exam fortnight — collecting marks from every subject teacher, tabulating, checking arithmetic, generating report cards
• March — consolidating a year of attendance out of twenty registers for the annual return
• Any day a parent asks a question the office cannot answer without walking to three different rooms

That is the honest list, and it is what to buy against. Anything that does not attack one of those five is a feature you are paying for out of politeness.

Fees: the one thing that must not stay a spreadsheet

If a school changes exactly one thing, change this one. Fees are the only area where a spreadsheet error stops being an inconvenience and becomes a scandal — with a parent who is certain they paid, a receipt book with a page torn out, and nothing in between to settle the matter.

What fees actually need is more structured than a spreadsheet can hold honestly: fee structures built from components like tuition, transport, laboratory and examination; monthly, quarterly or annual cycles; discounts, late fees and adjustments as recorded, approved entries rather than as someone typing over a cell; and receipts numbered in an unbroken series. Gap-free numbering is the entire audit story. A receipt book with missing pages cannot be defended, and a spreadsheet where anyone can insert a row cannot either.

Concessions deserve special attention because they are where trust is lost. Staff wards, sibling discounts, scholarships, RTE seats — in a spreadsheet these appear as a smaller number in a cell, with no record of who approved it or why. Recorded properly, each is a named discount with an approver against it, which protects the accounts clerk as much as it protects the school.

One tax note, because it separates two readers of this article. Services provided by an educational institution to its own students, faculty and staff sit under the education exemption in the GST notifications. A private coaching centre generally does not get that treatment and charges tax on its fees. If you run coaching rather than a school, your fee software has to produce proper tax invoices — and if you run both under one roof, confirm your exact position with your CA before assuming either.

Attendance and results: worth automating, or not

Attendance on its own is the most over-sold module in school software. If the only consumer of your attendance data is an annual return, digitising it buys you a tidier March and very little else. Say that out loud in the demo and watch how the vendor responds.

It becomes genuinely valuable the moment attendance feeds something. A percentage per student computed against the sessions that batch actually held rather than against calendar days. A parent who hears about three consecutive absences in the same week they happen. A risk view that combines attendance with fee delays and falling marks, so the school notices a child drifting before the fee ledger notices it in October. That combination is worth real money; a digital register by itself is a nicer-looking cupboard.

Results are the opposite case — the value is immediate and mostly arithmetic. Bulk marks entry against a defined marks scheme, ranks and percentiles computed rather than calculated by hand, and marksheets published without anyone rebuilding a formula sheet, removes a genuine fortnight of work per exam cycle and a class of errors that are embarrassing to correct after report cards have gone home.

The caveat is unavoidable: teachers must enter marks into the system directly. If they write on paper first and someone types it afterwards, you have added a step rather than removed one, and within two exam cycles the school will be back on the old sheet.

Parent-facing features, and the ones parents ignore

Parents care about a short list: what their child scored, whether their child was present, what is owed and what was paid, and when they can speak to a teacher. That is close to all of it.

What they reliably ignore is anything that requires installing an app and remembering a password. A school portal that a parent logs into twice a year is not engagement, it is a support burden — and a father driving an auto or a mother on a factory shift is not going to reset a password to find out a mark. The working rule: never make a parent install something to see a number. Send it to them.

Meetings are the exception worth investing in, because they are where anything actually gets resolved. A parent-teacher meeting scheduled properly — in person, by video or by phone, with an agenda and action items recorded against it — is worth more than any notification feed. The video and phone options are not a downgrade; for a parent working in another city, they are the only version of the meeting that will ever happen.

Do not buy school software for the parent-app demo. Buy it for the five things your office does badly every single month.

The chain-first ERP: what you will pay for and never use

Most school ERPs were built for groups, and it shows in what they lead with. On a single campus, this is the list you will be sold and will not switch on.

• Multi-campus consolidation, inter-branch transfers and group-level dashboards
• Hostel and mess management, if you have no hostel
• Vehicle telematics and live GPS, when what you needed was a list of which child gets off at which stop
• A full payroll engine sized for a thousand staff
• Store and inventory for lab consumables
• A biometric hardware stack with its own maintenance contract
• An admissions CRM with lead scoring, for a school that admits by walk-in and word of mouth

The cost of these is not only the licence. It is the three-month implementation, the consultant, the training days your teachers spend in a computer lab, and a per-student price built around a chain's economics. A single campus should be live in a week and self-sufficient by the end of the month.

So make one question decisive in every demo: which modules can I switch off, and does my price change when I do? A product organised as modules you turn on lets a small school buy the four it runs and ignore the rest. A product that only sells as one bundle has told you who it was built for.

A staged rollout across one academic year

The schools that fail at this go live with everything in the middle of a term. One new thing per term, each owned by a named person, is slower on paper and finishes far sooner in practice.

• Before the session opens: bulk-import students and guardians with row-level validation, set up batches, and build the year's fee structures. Go live on fees from day one of the session — mid-year fee migration is genuinely painful and worth waiting to avoid.
• Weeks four to six: attendance sessions per batch. One routine, teachers only, nothing else changes.
• Term two: one full exam cycle in the system — marks entry, ranks, marksheets. Run it alongside the old sheet for that one exam, compare, then retire the sheet on a fixed date.
• Term three: parent meetings and the risk view, once a term of attendance, fee and marks history exists to build on. A risk score computed from two weeks of data is noise, and teachers will judge the whole system by their first impression of it.
• Next session: library, reports, and whatever is genuinely left.

Migrate the minimum: students, guardians, batches, this year's fee structure, and the current outstanding. Do not type in five years of historical marks. Nobody has ever needed them badly enough to justify the fortnight, and the history rebuilds itself one exam at a time.

A single-campus school does not need a system that could run a group of colleges. It needs fees that cannot be argued with, attendance that feeds something useful, results that tabulate themselves, and a way to reach parents that does not depend on an app they will never open. Everything past that is somebody else's requirement, priced into your quote.