Rollout walkthrough
A day school moving off spreadsheets
How a first platform rollout sequences for a school whose current system is spreadsheets, and where the real work is data rather than software.
What this is, and what it is not
This is a composite, not a customer story. It describes how a rollout typically goes for this kind of school, drawn from implementation experience across several. No school is named, no figures are attributed to a customer, and nothing here is presented as a measured outcome. Named case studies appear on this page only with written consent and stated measurement methods, and none are published yet.
The school described: Around 600 students, single campus, one national curriculum, records spread across spreadsheets and a filing cabinet.
Starting position
What brings a school like this to look for a platform
The presenting complaint and the underlying condition are usually different things, and the difference decides the sequence.
Nobody disputes that the spreadsheets work. They do work, which is why they have survived. What has stopped working is that three people maintain overlapping copies, the enrolment figure the office quotes differs from the one finance quotes, and answering a parent question about a fee balance takes a phone call to a specific person who may be teaching.
The presenting complaint is usually reporting season. Report cards are assembled by hand from marks collected over email, the process takes two weeks of someone’s evenings, and a single correction after publication means reissuing a document. That pain is annual, acute, and remembered, which is why it is often the first thing a school asks to fix — and usually the last thing that should move.
The underlying condition is that the student roster has never been reconciled against a single authoritative source. Every downstream problem is a symptom of that, and every downstream fix will inherit it.
Sequence
What moves when, and why that order
Each phase states the reason it cannot sensibly come earlier or later. The reason matters more than the phase, because your school will need a different order for at least one of them.
- Before signatureStep 1
Data cleaning
Deduplicate students, deduplicate guardians, establish payer-of-record per child, and agree fee opening balances in writing with finance. Expect this to take longer than anyone estimates and to surface disagreements that predate the software decision.
Why here: This work does not depend on a vendor, it is the longest-lead item, and every later phase inherits its quality. A school that does this before signature has a calm rollout; a school that defers it has a data-cleaning emergency in go-live week with a live deadline attached.
- Term oneStep 2
Foundation and fees
Academic year, terms, class structure, and academic calendar first. Then the student and guardian import, verified against the school’s own enrolment count. Then fee structure, discount rules, gateway onboarding, and one full test cycle from invoice to reconciliation.
Why here: The calendar decides what a working day is, and attendance percentages, fee schedules, and timetables all read from it. Getting it wrong means recalculating everything built on top. Fees follow because the ledger attaches to the records the import created.
- Term twoStep 3
Timetable, attendance, portal
Build the timetable, verify that registers generated from it are accurate, train teachers on register taking, and only then open the family portal.
Why here: Registers come from the timetable, so the timetable has to be right first. The portal comes last because it is the first thing families judge, and a portal showing a wrong fee balance or a wrong attendance figure costs more trust than it earns.
- Term threeStep 4
Assessment and reporting
Configure assessment structure and grading, verify aggregate calculation by hand for a few students, generate a full cohort of report cards rather than a sample, and test correcting a published result.
Why here: Reporting aggregates everything above it. Moving it first — which is what the presenting complaint asks for — means aggregating data that is not yet trustworthy, and producing report cards that are wrong in a way that is hard to explain to parents.
Friction
What reliably goes wrong
Not risks in the abstract. These are the things that recur, so they can be planned for rather than discovered.
The enrolment count does not match
The office figure and the finance figure differ, usually by single digits, usually because of students who left without being marked as having left. Resolve this before proceeding rather than picking one. It is the cheapest reconciliation you will ever do and the most expensive to defer.
Teachers keep private mark lists
They will continue to, and they are not wrong to, until registers and marks entry are demonstrably accurate and quick. Time a real class set of marks entry with a teacher who has not used the system. If it is slower than their spreadsheet, the spreadsheet wins and no amount of policy changes that.
Nobody owns the project
The strongest single predictor of a difficult rollout is a project owned by "the office" rather than a named person with allocated time. Two half-days a week for a term is a realistic figure. Zero allocated hours with an expectation of goodwill is the common one.
The old process never switches off
Set the date the previous process stops when you plan each phase, not afterwards. A parallel period without an end date becomes permanent, staff split across both systems, and each one is now half-right.
Limits
What a platform does not fix for this school
Stated because the alternative is a school discovering it in term two.
- It does not resolve which of two near-identical guardian records is the correct one. Import validation flags the duplicate; the judgement stays with the school.
- It does not decide your fee policy. If the discount rules are hard to explain in a meeting, they will be hard to configure, and configuring them does not make them coherent.
- It does not remove the reporting cycle, only the manual assembly. Someone still has to decide what the report card says.
- It does not fix staffing. A school short of office capacity is still short of it, though it spends less of that capacity on re-entering the same figure.
A day school moving off spreadsheets: common questions
What schools in this position ask before committing to a sequence.
Can we do this in one term instead of three?+
Foundation and one workflow, yes. All four phases in a term, realistically no — not because configuration takes long, but because each phase depends on the previous one being verified, and verification happens at the speed of a school term rather than a project plan. Schools that compress it usually spend term two undoing term one.
Reporting is our worst problem. Why is it last?+
Because it aggregates attendance, assessment, and enrolment data. Moving it first produces report cards computed from data that is not yet correct, which is worse than the manual process it replaced — the manual process at least had a human checking each one.
What if our spreadsheets are actually well maintained?+
Then your rollout will be significantly easier than this scenario and the data-cleaning stage may take days rather than weeks. Run the counts anyway. Well-maintained spreadsheets still routinely contain duplicate guardians, because the duplicate is created by two children being enrolled at different times.
How much staff time should we budget?+
A named owner at roughly two half-days a week through the first two terms, plus a short training session per teacher for register taking and marks entry, plus the registrar’s time during the admissions cutover. The owner’s time is the part most often left out of the plan and most often the constraint.
Go deeper
Related planning material
Describe your own starting position
Tell us what your data looks like and which term you need to be live in. We will tell you what is realistic, including when the answer is not this year.
Already using Schoolyi? Sign in