Before you start: three decisions
Do not begin week one until these three are settled, because each one changes the whole plan:
- Who owns the project, by name? A project without a named owner stops at the first disagreement.
- What are the pickup points? One gate for every grade defeats the best system.
- Which class will you pilot? Pick a mid-sized class — neither the smallest nor the largest.
Week one — measure before changing
Nothing is switched on this week. You only measure. This is the week most commonly skipped, and skipping it destroys your ability to prove any improvement later.
| Metric | How to measure it |
|---|---|
| Longest queue | Count waiting vehicles every five minutes through the window |
| Peak minute | The minute at which the queue reaches its maximum |
| Average handover time | From guardian announcement to student released, for a daily sample |
| Yard clearance time | From first call to last handover |
| Exceptions | Count of alternate-collector cases and phone calls per week |
Week two — write the policies
This is where the project is actually decided. Three documents, each no longer than a page, approved with an administrative signature:
- Escort policy: who walks the student from class to the pickup point? When do they move after the name appears? What does the teacher do if the student does not move?
- Collector policy: how are authorised collectors registered? How is temporary authorisation granted, and who approves it? How are custody restrictions recorded?
- Outage policy: what happens if the internet drops or the screens fail? Who declares the switch to the manual path? How are records reconciled once service returns?
Do not move to week three until all three documents are signed and posted in the reception room. A policy that exists only in people's heads is not executed under pressure.
Week three — pilot one class
Run the system for the chosen class only, with the manual path operating in parallel. This week has three goals, and "the system works" is not one of them:
- Tune the geofence radius. Start from an estimate and adjust daily: if guardians arrive before students are ready, widen it; if students stand waiting, narrow it.
- Measure real preparation time. How long does a student take from name-on-screen to arriving at the pickup point? Everything else is tuned against this number.
- Inventory the failures. Log every case that did not go as planned — without exception and without excuses.
Week four — scale and stabilise
Expand to a full grade, not the whole school. Focus on three things most people miss when scaling:
- Onboard guardians in batches. Hold a scheduled session to help with installation and location permission. A guardian who fails to install falls back to the queue and slows everyone behind them.
- A backup for every role. Your reception staff member will be absent one day. If the system stops when they do, you implemented a person, not a system.
- Re-measure. Capture the week-one metrics again, the same way, and compare. That comparison is your case for continuing.
Checklist before going school-wide
- All three documents signed and posted where the work happens.
- Every authorised collector for the grade registered with documented consent.
- Custody restrictions entered and reviewed with administration.
- Screens installed at every pickup point and tested in daylight, not only indoors.
- A trained backup for every key role.
- Post-rollout metrics captured and compared to the baseline.
- The outage policy rehearsed in practice at least once.
Mistakes that recur in every stalled rollout
- Going school-wide at once "to save time" — which costs months later.
- Skipping the measurement week, making impact impossible to prove and renewal hard to justify.
- Leaving the geofence radius at its default with no field tuning.
- Onboarding guardians with a single message instead of a real help session.
- Retiring the manual path early out of confidence in the system.
Frequently asked questions
How long does implementing a school calling system take?
Four weeks is enough to reach a full grade running stably: one week measuring, one writing policies, one piloting a class, and one scaling. Going school-wide after that depends on the pace of guardian onboarding.
Can it be implemented mid-year?
Yes, and it may be better: procedures are settled and guardians are used to the dismissal pattern, so the effect of the change shows clearly against a known baseline.
Which step is skipped most often?
The measurement week. Skipping it feels like saving time, but it robs you of any proof of impact — turning the conversation with administration into competing impressions with no way to settle them.
Should the manual system be retired after rollout?
No. The manual path remains permanently as a fallback for outages and exceptions. The only requirement is that it be recorded in the same audit log, so no gap opens.
Who should lead the project inside the school?
Someone from operational administration with authority to change procedures — not the IT lead. Most project decisions are operational (pickup points, escort policy), not technical.
Read next: