Every school ICT manager already knows this instinctively, but it rarely makes it into a project plan: in a school, the calendar is the binding constraint. Not budget, not scope, not vendor lead times. The calendar.
Commercial IT projects can absorb a slipped date. A school cannot start term with a half-migrated network. Planning backwards from the calendar rather than forwards from the purchase order is the single change that most improves how school infrastructure projects run.
The four windows, and what each is actually good for
The summer break
The only genuinely long window, and the only one suitable for anything that touches core identity or the network backbone. Domain and directory work, core switching replacement, server and storage migration, wireless overhauls across multiple buildings, and anything requiring a rolling outage measured in days.
It is also the window everybody wants, which is why it needs to be booked with vendors and installers well ahead of the final term.
The mid-year break
Long enough for substantial work but not for anything you cannot reverse. Good for staged wireless rollouts, firewall replacement with a tested rollback, endpoint fleet refresh in a building or two, and backup platform changes.
The short breaks
Roughly two weeks, and realistically about eight usable working days once leave is accounted for. Suitable for contained, reversible work: a single-site wireless refresh, a lab rebuild, a firmware campaign, licence migrations, pilot deployments.
The failure mode here is optimism. Work that needs eleven days does not fit into eight simply because the purchase order was signed.
Term time
Monitoring, documentation, policy work, procurement, planning, pilots with a small volunteer group, and anything that can be genuinely rolled back inside a period bell. Not core infrastructure.
The lead times nobody plans for
The break is not the project. The break is the cutover. The work that determines whether the cutover succeeds happens in the weeks before it, and each of these has bitten schools we have worked with:
- Hardware lead times. Enterprise networking and server hardware is routinely quoted at two to six weeks, and longer when a vendor is between product generations. Order timing, not install timing, is the real deadline.
- Structured cabling and electrical. If new access points or racks need power or fibre, a separate trade is involved, with its own booking calendar and its own compliance sign-off.
- Board and finance approval cycles. Capital approval often runs on a meeting schedule that does not align with when equipment needs ordering.
- Vendor and distributor shutdowns. The December to January period, which is precisely your biggest window, is also when many suppliers and their logistics partners are closed.
- Licensing and subscription anniversaries. Renewal dates that fall mid-project create awkward decisions about whether to renew something you are about to replace.
Working backwards
A useful discipline is to write the plan in reverse. Start from the first day students return and work back:
- Students return. Everything must be stable and staffed for a heavy support day.
- Staff return, a few days earlier. This is your real deadline. Staff arriving to a broken environment is how a technically successful project gets remembered as a failure.
- Buffer. At minimum two working days of contingency before staff return. Non-negotiable, and the first thing people delete when the plan gets tight.
- Testing and validation. Not just does it power on, but does printing work, does the sign-in flow work on a student device, does the curriculum application still authenticate.
- Cutover work. The actual installation and migration.
- Pre-staging. Configuration built and tested in advance so the break is spent installing rather than designing.
- Delivery and goods inward. Equipment physically on site, unboxed and checked, before the break begins.
- Order placed. Working back from delivery by the quoted lead time, plus a margin.
- Approval and purchase order. Working back to the board or finance meeting that authorises it.
Run that backwards for a summer core network replacement and the approval date usually lands somewhere in the middle of term three. Schools that discover this in November end up either rushing or waiting a full year.
Two things worth building into every school project
A tested rollback. Not a theoretical one. If the new firewall misbehaves on the first Monday, what exactly gets reverted, by whom, and how long does it take? Write it down before the break, not during it.
A first-week support plan. The week students return generates more tickets than any other week of the year, and a recent change concentrates them. Plan for onsite presence rather than remote-only coverage.
The short version
Decide what you want live for the start of the school year, then work backwards through testing, cutover, staging, delivery, ordering and approval. Whatever date that produces is your actual deadline, and it will be earlier than feels comfortable.
If you are mapping out infrastructure work for the coming year and want a second opinion on whether it fits the windows available, we are happy to look at it with you.
