A timesheet answers three questions at once: how long did this person work, on what, and at whose cost? The daily time card answers the first question for one day; the timesheet answers all three for a week or two, and it is the document that payroll, billing, and project accounting all read. Its structure — days across the top, projects or tasks down the side, totals in both directions — has been stable for a century, from paper forms to software, because the questions it answers have not changed. What has changed is the stakes: a timesheet that once settled a weekly wage now also settles client invoices, grant compliance, and utilisation rates, and an error in it propagates into all three.

Anatomy of a timesheet

A standard timesheet has a header (worker, period, approver), a grid (days across, projects or tasks down, with regular hours, overtime hours, and leave separated), and a footer (totals, signature, approval). Each cell is the hours worked on that task on that day, usually to the nearest quarter-hour or tenth of an hour, and each row sums to the task’s weekly total while each column sums to the day’s total. The grand total — rows and columns agreeing — is the cross-foot, and it is the first check a reviewer performs: a timesheet whose rows and columns disagree has an arithmetic error, and the error is found by re-adding, not by trusting either total.

The categories down the side are the timesheet’s real content. For an employee, they are usually regular time, overtime, sick leave, annual leave, and holidays. For a consultant or agency worker, they are the projects and clients, with billable and non-billable time separated — administration, training, business development, and bench time are real work that no client pays for, and recording them honestly is what makes utilisation figures meaningful instead of flattering. For grant-funded and government-contract work, the categories are the funding codes, and the timesheet is the document an auditor reads first, because labour is usually the largest cost and the easiest to misstate.

Billable time, utilisation, and the Friday reconstruction problem

Billable hours are the hours charged to a client at an agreed rate; non-billable hours are everything else. Utilisation is billable hours divided by available hours (usually 40 a week, less leave), and a utilisation of 75% — 30 billable hours of a 40-hour week — is a strong figure in professional services, not a weak one, because the remaining quarter is the administration, training and selling that keeps the billable work coming. A utilisation target of 100% is a target that guarantees either falsified timesheets or burnout, and usually both in sequence.

The enemy of accurate timesheets is memory. Hours reconstructed on Friday from a week’s recollection systematically understate short, fragmented tasks (the 15-minute calls, the email triage) and overstate long, continuous ones, and the errors do not cancel — the total drifts, and the project split drifts more. The fix is contemporaneous recording: enter the hours as the work happens, or at the end of each day, while the day is still retrievable. Software that prompts at day’s end, timers that run during the task, and calendar integration that suggests entries from the day’s meetings all serve the same purpose — moving the recording closer to the work, where the memory is still accurate.

Overtime and leave on the timesheet

Overtime on a timesheet is the weekly excess past 40 hours, at 1.5 times the regular rate, and it belongs in its own row or column — not folded into the regular hours, where it becomes invisible to payroll and to the project’s cost. The regular rate for the premium includes non-discretionary bonuses and differentials, so the overtime row’s rate is not always the hourly rate, and the timesheet’s job is to supply the hours while payroll supplies the rate. Leave — sick, annual, holiday — is recorded as hours in its own category, paid at the regular rate, and excluded from the overtime calculation: a week with 32 hours worked and 8 hours of holiday is 40 hours paid and zero hours of overtime, because overtime counts hours worked, not hours paid.

The timesheet calculation on this site sums the week with lunch breaks deducted where the card records them, and the time-card calculation supplies the daily totals the timesheet sums. The two should reconcile — the timesheet’s week is the sum of the cards’ days — and a difference between them is an error in one of the two, usually a break deducted twice or a day entered in both.

Approval, corrections, and the audit trail

Approval is the control that makes a timesheet trustworthy. The worker attests that the record is complete and correct; the manager approves it, confirming the work was authorised and the coding is right; both attestations are dated, and the approved timesheet is locked against further editing without a new approval cycle. A correction after approval reopens the cycle — the correction, who made it, and why are recorded, and the approval is re-obtained. A system that lets anyone edit an approved timesheet silently has no control at all, whatever its interface claims.

Deadlines are part of the control: timesheets submitted late delay payroll, delay invoicing, and are less accurate, because the memory they rely on is older. The usual discipline is submission by a fixed time on the first working day after the period ends, approval within a day, and payroll and invoicing run from the approved set only. Exceptions — a late timesheet, a corrected timesheet, a timesheet approved by someone other than the usual approver — are logged, because in an audit the exceptions are the sample.

Timesheets for freelancers: the invoice’s evidence

For a freelancer, the timesheet is the invoice’s evidence: the hours, the dates, the tasks, and the rate, in a form the client can verify against their own records of meetings and deliveries. A freelance timesheet should record the task in enough detail that the client recognises the work — “project work, 6 hours” is not evidence, “drafted sections 3–5 of the report, 6 hours” is — and it should separate billable hours from the unbilled time (proposals, administration, revisions within scope) that the rate must absorb. Retainers change the shape but not the discipline: the timesheet shows the retainer’s hours being consumed, and the client sees the balance, which is what prevents the end-of-month argument about whether the retainer was used up.

Frequently asked questions

What is a timesheet?

A record of hours worked over a period — usually a week or two — broken down by day and by category (project, task, regular time, overtime, leave). It is the document payroll builds the paycheck from, billing builds the invoice from, and project accounting builds utilisation from. Each row sums to a task total, each column to a day total, and the two must agree.

What is the difference between a timesheet and a time card?

A time card records one worker’s day — start, end, breaks, total. A timesheet records a period — a week or two — by day and by project or category. The timesheet’s week is the sum of the cards’ days, and the two should reconcile. In office and project work the timesheet is often the primary record, filled in by the worker rather than stamped by a device.

How do you calculate total hours on a timesheet?

Add the hours for each day, deducting unpaid breaks, and sum the days for the week. Daily totals should be entered to the nearest quarter-hour or tenth of an hour, and the row totals (by task) and column totals (by day) must cross-foot to the same grand total. Overtime is the weekly excess past 40 hours worked, in its own category.

What are billable hours?

Hours charged to a client at an agreed rate, as opposed to non-billable hours — administration, training, business development, bench time — that no client pays for. Utilisation is billable hours as a share of available hours, and 75% is a strong figure in professional services, because the remaining quarter is the work that keeps the billable work coming.

Who approves a timesheet?

The worker attests it and the manager approves it, confirming the work was authorised and the coding is right, with both attestations dated. The approved timesheet is locked against silent editing; a correction reopens the approval cycle and is recorded with who made it and why. Payroll and invoicing run from the approved set only.

How often should timesheets be filled in?

Daily, as the work happens or at the end of each day. Hours reconstructed on Friday from memory systematically understate short tasks and overstate long ones, and the errors do not cancel. Contemporaneous recording — timers, calendar integration, an end-of-day prompt — is the single most effective accuracy measure.