Time and attendance in care: from clock-on to a timesheet payroll can trust

If you run rosters for an NDIS, aged-care or community provider, you already know the roster and the timesheet are two different documents. This guide is for coordinators and finance officers who need to understand the gap between what was planned and what actually happened — and how to close that gap without reading every shift by eye.

A roster is a plan: this worker, this client, this time, this address. A timesheet is a record: this worker clocked on at this time, from this place, and clocked off at that time, from that place. Payroll and billing should run off the timesheet, not the roster — because the roster can be wrong (a worker runs late, a shift extends, a visit happens somewhere other than the usual address) and the business shouldn’t pay or claim for a plan, it should pay and claim for what happened.

This is the core idea behind any attendance roster software: the roster and the timesheet need to talk to each other, but they’re not the same thing, and the office needs a clean way to see where they differ.

A worker starting a shift from their phone should produce three pieces of information without a phone call to the office:

• The time — when the shift actually started and finished, not when it was rostered to.
• Where the phone was — a position captured at clock-on and another at clock-off.
• Task and note confirmation — the worker should be prompted to confirm what they did and, where the shift requires it, add a note, before the clock-off is accepted.

None of this is about watching a worker move through their day. It’s two points in time — on and off — each tied to a position, so the office has something to compare against the plan afterwards. A system that gates the clock-off behind the right tasks and notes also saves the coordinator a second job: chasing missing documentation after the fact.

Every visit has an address — the client’s home, a day program, a community outing. When a worker clocks on or off, the position their phone recorded can be compared with that address. If the two line up, within a sensible distance, there’s nothing to look at. If they don’t — the clock happened well away from where the visit was supposed to take place — that’s a timesheet exception, not an automatic rejection.

This matters because the comparison is a prompt for a person, not a verdict. A clock a few hundred metres from the front door could mean the worker parked down the street, or it could mean something worth a quick check. Either way, the right design flags the exception on the timesheet and leaves the decision with the office — it doesn’t block the shift, and it doesn’t pretend to know what happened on the ground. Treat it as one input into an approval decision, alongside everything else on the timesheet.

If you also want to see this kind of thing as it happens during the shift, rather than after, that’s a different problem — see our guide on real-time visibility in care operations for what the office can usefully watch live versus what’s better left to the end-of-shift review.

Not every timesheet needs a human eye. Most won’t. The ones that should get a second look before approval typically fall into a short list:

• A clock outside the expected distance from the address — worth a glance, not necessarily a problem.
• A shift that ran meaningfully longer (or shorter) than rostered — did the visit genuinely take longer, or did someone forget to clock off?
• Travel between clients on the same day — common in community and in-home work, and worth checking that travel time is being recorded against the right shift rather than absorbed into one or the other.
• A shift with no matching roster entry at all — a worker covering an ad hoc visit, or a data entry gap.

Everything else — a clock-on and clock-off that line up with the roster, within a sensible window, at the right address — doesn’t need a person. That’s the point of separating exceptions from matches: the office’s time goes to the handful of shifts that actually need judgement, not to re-checking every shift that already matches the plan.

Once exceptions are flagged, the rest of the week’s timesheets should move through approval as a batch, not one click at a time. A timesheet queue that lets you select everything that matches — right time, right place, right tasks completed — and approve it in one action is what turns a day of checking into twenty minutes. The exceptions sit outside that batch until someone has looked at them and either approved, adjusted or queried the time with the worker.

This is also where paid travel time belongs: if a worker travels between clients and that travel is paid, it should be recorded against the shift so it moves through the same approval step as the rest of the timesheet, rather than being handled as a separate reimbursement process after the fact.

Once a timesheet is approved, that’s the record of truth for two separate processes that both need to agree with each other: the pay run, and the funding claim.

For pay, the approved hours are what get costed against the applicable pay conditions — loadings, allowances, the lot. We’ve written separately about how that costing actually works under SCHADS-style award logic in SCHADS for rostering: the parts that actually bite, which is worth reading alongside this guide if you’re setting up bands and loadings for the first time.

For billing, the approved hours are what should be claimed — not the rostered hours. If a shift ran over, or started late, or didn’t happen at the client’s usual address, the claim should reflect what was approved on the timesheet, because that’s the version of events that will need to stand up if a funder or auditor asks. Running payroll and claiming off two different numbers — the roster for one, the timesheet for the other — is a quiet but common source of reconciliation headaches down the track.

Corella’s rostering board carries the plan through to a clock-on and clock-off from the worker’s phone, with task and note gates before a shift can be closed out. Where the visit’s address has been placed on the map, each clock-on and clock-off is compared with it, one point each time. A clock outside the distance you set (1 km unless you change it) is flagged on the timesheet with how far away it was, and that timesheet waits for the office instead of being approved automatically. Paid travel time is recorded on the shift itself, and the timesheet queue supports bulk approval for everything that matches, and approved hours go out as a payroll CSV for your payroll software or, where your Xero is connected, to Xero Payroll as draft timesheets.