Shift Planner & Roster

What this tool does

CM8-268 is a working shift roster. One row is one assignment — one person, on one date, on one shift. From those rows it builds the week: the roster view shows headcount by shift for each day and flags any shift below the target you set; the coverage table does the same arithmetic in numbers; the hours chart shows who is carrying the week; and the printed report is a roster the team can actually read.

Everything runs inside this single file. There is no account, no upload and no network request of any kind, so names, absences and working patterns never leave the computer you are using — which matters, because a roster is personal data about identifiable people.

Not the manpower planner, not the leave planner

Three tools sound alike and do different jobs. The Manpower Planner does capacity arithmetic — how many people a workload needs, in aggregate, before anyone is named. The Employee Leave & Absence Planner tracks entitlements — who has how much leave left, and whether requests overlap. This tool is the operational layer underneath both: who works which shift on which day, checked against the headcount each shift needs. Use the manpower planner to decide the targets, the leave planner to manage the balances, and this roster to put names against days. A person shown as "on leave" here is a rostering fact for coverage purposes; their remaining entitlement lives in the leave planner.

Building a week

Requirements come first. On the Settings tab, set the headcount each shift needs on a normal day — early, late and night. Those three numbers are the whole standard the roster is judged against, so agree them before arguing about names: a coverage gap against a wrong target is noise, and against an agreed one it is a decision waiting to be made.

Then add people, one assignment at a time: date, person, team, shift, hours. Record the non-working entries too — rostered off, on leave, sick, training. They contribute nothing to coverage, but they are the difference between "not rostered" and "deliberately off", and they are what makes the absence view honest.

The coverage table then shows the truth: for every rostered day, each of the three target shifts with required, rostered and the gap. Days with no assignments at all are treated as not yet planned rather than short — an empty Thursday is a to-do, not a crisis.

Reading the coverage table

Each row is one shift on one day. Required is your Settings target, Rostered is a headcount of assignments on that shift, and Gap is the shortfall — never negative, because an extra person is not a problem the table needs to flag. The status column says OK or SHORT and nothing subtler, deliberately: a roster check you can read at a glance gets read; one that needs interpreting does not.

Day-shift and training assignments do not count toward the early, late or night targets. If your operation runs a single day shift, set the early target to your day headcount and roster everyone as early — or set the targets you do not use to zero.

Shift patterns in one paragraph

Fixed patterns — the same people always on nights — are simple to plan and let bodies adapt to one rhythm, but they concentrate the unsocial hours on the same few names. Rotating patterns share the burden but cost more to plan and more to adapt to. Where teams do rotate through nights, common practice is to rotate forward — early, then late, then night — and to keep runs of consecutive nights short, because most people find that direction easier to live with. That is working custom, not medical advice; if you run sustained night work, take proper occupational health guidance.

Hours and overtime discipline

The hours chart totals each person's planned hours across the latest week and turns red above the standard weekly hours from Settings. One red bar in a heavy week is life; the same red bar every week is a staffing decision being made by default. The overtime checkbox marks an assignment as beyond someone's contracted hours — the tile counts those shifts, their hours and the people carrying them.

Read chronic overtime as a signal, not a fix: if the roster only balances when the same two people are flagged for overtime every week, the honest reading is that the shift is understaffed by roughly that many hours, and the overtime budget is quietly paying for a vacancy.

Absence patterns — kindly, but honestly

The four-week view shows leave and sick days per person. Booked leave clustering in school holidays is a plan working. Sickness is different: it is information about a person, a shift or a workload, and the chart can only tell you that a pattern exists — never why. A run of sick days after night blocks might be the pattern, the workload, or something at home. The right response to a pattern is a private, supportive conversation, not a chart waved across a table; the wrong response teaches people to come in ill, which costs more than the absence did.

Planned, confirmed, worked, changed

An assignment starts as planned — a draft only the planner has seen. When the roster is published to the team, mark rows confirmed; the unconfirmed tile counts what is still in draft so you never publish half a week by accident. After the day, mark rows worked to record that they happened as rostered — sickness discovered on the day is recorded the same way, by changing the shift to sick and the status to worked. Changed marks a row altered after publication — a swap, an extension, a late cancellation. Those rows are your audit trail: if the printed roster and the worked week disagree, the changed rows are the explanation, and how many of them there are each week is a fair measure of how much notice your team really gets.

Skills coverage on shift

Some cover is not about headcount: a shift can be four strong and still have nobody who may drive the forklift or act as first aider. The skills field sits on the assignment, not on the person, because what matters is what is covered on that shift — the qualified person on leave covers nothing. Note the roles that must never be absent, and scan the roster table's skills column before confirming the week. If a shift shows no first aider, that is a gap the headcount arithmetic cannot see.

Rostering as a fairness system

A roster is one of the most closely read documents in any workplace, and people read it for fairness before they read it for accuracy. Three things carry most of that weight: predictability — publish the roster far enough ahead that people can plan lives around it, and treat late changes as the exception the changed status makes visible; rotation equity — unsocial shifts, weekends and the overtime that pays well should circulate rather than settle on the same names, and the hours and shift-mix views will show you whether they do; and consistency — the person who always gets the awkward shift notices, even if nobody says so. Many places also have rules or agreements about advance notice and rest between shifts; whatever applies to you, the roster is where compliance becomes visible.

The formulas

"The latest week" everywhere means the 7 days ending on the most recent date in the register, so the figures follow the roster rather than the calendar. The absence rate's denominator is scheduled attendance — working assignments plus the sick entries that replaced them. Rostered-off and leave rows are excluded from both sides: someone on booked leave is not absent, and counting them would flatter or damn the rate depending on the season.

FAQ

Can I plan more than one week? Yes — add assignments for any dates. The tiles, charts and coverage table always read the latest 7 days with data; the register and CSV hold everything, and the date filter lets you look at any earlier week.

We run 12-hour shifts, not three shifts. Use early for days and night for nights, set the hours to 12, and set the late target to zero. The arithmetic does not care how long a shift is — only headcount against target, and hours against the weekly standard.

Why doesn't an empty day count as a coverage gap? Because "not yet rostered" and "rostered short" need different responses. The roster chart still shows empty days as "not rostered" so they cannot hide; they just do not inflate the gap count while the week is half-built.

How do I record a swap? Edit both rows — the person coming in and the person now off — and set both to changed, with a word in the note. Two changed rows that reference each other are a complete record of the swap.

Does the hours figure feed payroll? No. These are planned hours; payroll needs worked hours. Marking rows worked (and correcting hours where the day differed) closes most of that distance, but keep your time records as the source of truth.

Saving your work

Assignments, settings and the report header are written to this browser's local storage as you type, and the toolbar shows the time of the last save. That storage belongs to one browser on one computer: another browser, a private window, a second machine or a clean-up tool that clears site data will not have it.

Treat Export .json as the real save — one file containing everything, which Import .json restores anywhere. Export CSV gives you the roster for spreadsheet work and includes every filtered record. Reset asks twice, then erases everything this tool has stored. There is no undo. A roster names people and their absences — treat exports as confidential.

Accuracy & disclaimer

The planner does coverage arithmetic and does it faithfully: people rostered against the headcount you set, hours against the weekly standard you set. Everything that matters sits underneath the arithmetic — whether the targets are right, whether the hours entered are real, and whether the pattern you have built is one people can sustainably work.

Local rules on maximum hours, rest between shifts, breaks, night-work limits and advance notice of rosters differ by country, by sector and by agreement, and this tool does not check any of them — applying them is yours to do. This is a planning and record-keeping aid, not employment-law advice and not a payroll system.