Home · Business · Operations & Logistics · Scheduling & rotas
As of 13 August 2026, AI can only partly schedule staff around customer demand.
This still needs a person who signs their name to it.
Can you do it?
15 minutesto a draft.
1 hourto something you’d act on.
Cost, all in£0
Skill neededpower-user
Who has to check ita colleague
What the alternative costsNo priced human or software alternative is provided in the available tool data.
If this goes wrong: busy periods are left understaffed or people are assigned unsuitable shifts, causing service failures, complaints or employment problems.
What to actually do
Hand it to a person
The route this page recommends
A person who owns the outcome does this end to end, worth it when the failure is dear.
Use a tool built for this
Second choiceDo it yourself
The distant thirdA chat interface, power-user skill, and roughly 1 hour until you can act on the result.
How to actually do it
- Open the demand report, till report or booking system and export demand by hour or another useful time period for the planning dates.
- Open the staff availability, leave and skills records and combine them with contracted hours, shift rules, opening hours and minimum coverage requirements.
- Remove unnecessary personal details, label each staff member with a role or ID, and paste the prepared tables into the prompt.
- Ask the chatbot to produce the draft schedule, coverage check, hours totals, uncovered periods and rule conflicts in separate tables.
- Compare every assigned shift with the original availability, leave, skills, contracted hours and supplied workplace rules, correcting the draft where it conflicts.
- Compare each time period's staffing level with the demand report and ask a colleague who understands the operation to check the forecast, fairness and practical coverage.
- Enter the agreed rota into the scheduling or payroll system and send it to staff only after the manager accepts the final version.
Prompt
Create a draft staff schedule around customer demand using the information below. Business and planning period: - Business type: [business type] - Location and time zone: [location] - Dates covered: [dates] - Opening hours: [opening hours] Demand: - Demand forecast or historical demand by [hour or other time period]: [paste table] - Minimum staff needed for each time period, if known: [paste table] - Any demand peaks, events or closures: [details] Staff: - Staff names or IDs: [paste list] - Skills, roles and maximum number of people needed in each role: [paste table] - Availability and unavailable times: [paste table] - Contracted hours, minimum or maximum shift lengths, rest requirements and other workplace rules: [paste the rules] - Preferences, leave and agreed restrictions: [paste details] Scheduling rules: 1. Use only the facts and rules supplied. Do not invent availability, qualifications, demand or legal requirements. 2. Prioritise required coverage and suitable skills, then distribute hours fairly and respect availability and supplied workplace rules. 3. Identify any period that cannot be covered rather than assigning someone who is unavailable or unqualified. 4. Produce a table showing date, shift start, shift end, role, assigned staff member and expected demand. 5. Add a coverage check for every time period, total hours by staff member, uncovered periods, rule conflicts and assumptions. 6. Give two alternative changes that would improve coverage, such as moving a shift or adding a person, but do not present an assumption as a fact. 7. Do not publish or send the rota. This is a draft for a manager to check against the source records and workplace rules.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- It cannot see live bookings, queue lengths, absences or changing demand unless you provide current data or connect a suitable system.
- It cannot decide whether a forecast is credible when demand is unusual, seasonal or affected by an event that is missing from the data.
- It can apply the workplace rules you provide, but it cannot reliably establish every employment, equality or working-time requirement for your situation.
- It cannot take responsibility for the effect of understaffing, unsuitable assignments or an unfair distribution of shifts.
- It does not replace a rota system that manages availability, notifications, holiday records, attendance and last-minute changes.
What caps this at PARTLY: judgement under ambiguity, verification cost and stakes of error.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 1 |
| Verification | 1 |
| Liability | 1 |
| Effort delta | 2 |
| Total | 7 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can ChatGPT create a staff rota based on customer demand?
- Yes, it can create a draft rota from demand, availability, skills and coverage rules. You still need to check the draft against the source records, workplace rules and actual operating conditions before using it.
- What information does AI need to schedule staff around demand?
- Give it demand by time period, opening hours, minimum staffing levels, staff availability, skills, leave, contracted hours and the shift rules you use. It cannot fill gaps safely when those inputs are missing or out of date.
- Can AI predict how many staff I need for each shift?
- It can analyse historical demand or a forecast that you provide and turn it into suggested staffing levels. It cannot know whether an unusual event, booking pattern or operational problem will make that forecast wrong.
- Is an AI-generated staff rota safe to use?
- Use it as a draft, not as an automatic decision. A manager should check coverage, availability, skills, fairness and the workplace rules before publishing it, because the employer remains responsible if the rota causes problems.
Nearby answers
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.