Map demand, workload, roles and capacity signals in 15-minute blocks so a café can test small changes against comparable periods without treating a sales spike as a staffing answer.
New for 2026 practical guide
At a glance
Measure this
Workload Pressure per Role-Hour: Workload units recorded in a 15-minute interval divided by actual total role-hours, reviewed beside trigger minutes and matched-period context.Stop Staffing by Instinct: Build a 15-Minute Café Rush Map
Observe the work, not only the sales line
A 15-minute café rush map is a dated grid that puts POS demand, workload, people by actual role, constraints and short observations in the same place. It does not predict one correct staffing number. Its purpose is to help an owner or shift lead ask where the work exceeded the capacity of the roles and layout actually available, then test one small response against comparable periods.
POS timestamps do not show queue length, drinks in progress, cleaning work, stock-outs, machine faults, allergy conversations or team competence. Pair the export with a brief floor observation. Do not set a labour-saving target before mapping the current operation, and do not treat a difficult interval as proof that an individual failed.
1. Fix the map boundary and POS definition
Start with one service period, such as Monday–Friday 07:00–11:00 counter and collection service. State whether delivery, table orders, retail items, voids and refunds are included. Choose the POS timestamp closest to order acceptance—order created, payment completed or another documented field—and do not change that definition halfway through a comparison.
Use fixed, start-inclusive and end-exclusive intervals: 08:00–08:14:59, then 08:15–08:29:59. Export only business date, selected timestamp, de-duplicated order ID, channel, sales/refunds, item/quantity/modifier and relevant void/refund reason. Keep customer names, payment details and free-text notes out of the grid.
2. Turn menu mix into a transparent workload proxy
Order count is useful but incomplete. Add locally agreed workload units (WU) so a multi-item heated-food order is not treated as the same work as a packaged item. The weights below are illustrative prompts, not industry standards or production-time promises.
| Activity | Illustrative WU | Why it is visible |
|---|---|---|
| Grab-and-go retail | 0.5 | Till interaction and hand-over. |
| Batch coffee, tea or bottled drink | 1.0 | Pour/make and hand-over. |
| Espresso-based drink | 2.0 | Bar workflow and equipment demand. |
| Heated food | 2.0 | Food production and hand-over work. |
| Modifier | +0.5 | Extra communication/preparation. |
| Collection or delivery order | +0.5 | Labelling, checking and staging work. |
Workload units (WU) = sum of quantity sold × documented WU weight for each activity
Keep the first version simple. If observation shows a regular item creates more work than its weight suggests, change the documented weight for the next review cycle, not retrospectively. The goal is comparability, not false precision.
3. Make a POS/labour grid around actual roles
Record what people actually did in the interval, not merely their rota title. A colleague may be 0.5 till and 0.5 expo; do not count them as a full person in both roles. A useful role set is shift lead/floater, order point/till, bar, food/oven, expo/hand-over and runner/reset.
| Role | Peak responsibility | What to record |
|---|---|---|
| Lead/floater | Prioritises, handles exceptions and redeploys | Coverage fraction, intervention and constraint. |
| Till | Takes orders/payment and follows the agreed allergy prompt | Coverage and queue observation. |
| Bar | Produces drinks and resets the immediate station | Coverage and machine/ingredient constraint. |
| Food/oven | Produces food by the approved method | Coverage and prep/oven constraint. |
| Expo | Checks, calls, stages and hands over orders | Coverage and hand-over congestion. |
| Runner/reset | Replenishes, clears permitted obstructions and supports flow | Coverage and cleaning/stock interruption. |
Protect competence as well as a headcount. The FSA says staff need enough training to provide accurate allergen information, or must know to defer the order to a specified trained person. 1 Do not count an untrained colleague as equivalent cover for that role.
Use three operational states: stable, where hand-offs are moving; pressured, where a queue, work-in-progress point or hand-over builds across two checks; and constrained, where a capacity trigger, safety/food-control issue or essential resource failure needs an immediate response. The response may be redeployment, a clearer hand-off, future extra cover, a simplified offer or further investigation. It is not an instruction to remove labour.
4. Agree capacity triggers before the rush
A trigger is a measurable local condition that prompts a pre-agreed response. It is not a target staff must beat. Test a small number for a fortnight and revise them using the physical layout, equipment, menu, channels and current risk assessment.
| Trigger | Local definition example | Immediate response |
|---|---|---|
| Order-point build-up | Four parties waiting on two checks a minute apart | Lead supports till or opens trained/safe cover. |
| Production WIP | More than eight paid tickets at the constrained station on two checks | Protect the constraint; pause non-urgent side tasks. |
| Hand-over congestion | Marked hand-over spaces full or customers cannot distinguish orders | Assign expo/runner and restore staging discipline. |
| Essential resource | POS, oven, machine, key ingredient or approved stock unavailable | Use the documented contingency and communicate accurately. |
| Safety/food-control issue | Spill, blocked route, missed required check or allergy uncertainty | Make safe, clean/isolate/escalate and follow the site process. |
The numerical examples are not universal thresholds. HSE catering guidance highlights prompt removal of spills and clear space free from obstructions. 2 Do not sacrifice cleaning, food controls or allergy communication to keep a queue moving.
5. Map four to eight comparable peaks
Prepare the grid before service. At each 15-minute boundary, the lead records POS aggregates, role-hours, channel mix, state, trigger duration and one factual note. If recording cannot be done safely at the time, add it immediately afterwards and label it retrospective. A useful one-minute scan asks: how many parties are at the order point; how many tickets await the constrained station; are hand-over spaces full; are routes clear; is equipment/stock missing; and has a safety or food-control issue appeared?
Compare only matched periods: same day type, same interval and as similar an operating context as possible. Mark school holidays, weather, events, road closures, equipment outage, stock-out, menu change, absence and one-off bookings. Use the median of three to six eligible matched occurrences. With fewer than three, label the result insufficient comparison base rather than making a staffing decision.
| Scorecard measure | Current interval | Median eligible match | Interpretation prompt |
|---|---|---|---|
| Orders and WU | Did demand or product/channel mix differ? | ||
| Role-hours | Was actual role coverage comparable? | ||
| WU per role-hour | Signal to inspect, not a productivity verdict. | ||
| Trigger minutes | Where was capacity constrained? | ||
| Safety/food-control interventions | What process needed protection? | ||
| Constraint/action note | n/a | What is the smallest safe next test? |
For a controlled test, write one hypothesis: “On ordinary Tuesdays, assigning the trained floater to expo from 08:30–09:00 may reduce hand-over trigger duration; we will observe four eligible Tuesdays.” Do not claim a successful outcome from an anecdote.
Primary KPI: Workload Pressure per Role-Hour (WPRH).
WPRH = workload units in the 15-minute interval ÷ total actual role-hours in that interval
Four people each covering a full 15-minute interval equal 1.00 total role-hour; one person covering the full interval equals 0.25. Do not calculate WPRH when total role-hours are zero. Report it beside trigger minutes, context and the note. A higher value alone does not establish a cause or prove service speed, productivity or staffing need.
Printable worksheet
15-MINUTE CAFÉ RUSH MAP — POS/LABOUR GRID
Date: __________ Site: __________ Lead: __________ Service boundary: __________________
POS timestamp: order created / payment completed / other: __________
Context: NORMAL / SCHOOL-HOLIDAY / EVENT / WEATHER / OUTAGE / STOCK-OUT / OTHER: ________
BEFORE SERVICE
[ ] Time boundary and POS field confirmed [ ] Rush roles allocated
[ ] Allergen escalation route briefed [ ] Routes/hand-over space clear
[ ] Stock/equipment changes checked [ ] Breaks and cover planned
[ ] Trigger definitions and responses briefed
Interval | Orders | WU | Counter/collection/delivery | Till hrs | Bar hrs | Food hrs | Expo hrs | Runner/lead hrs | Total hrs | Queue/WIP | Trigger/minutes | Safety/food action | Constraint/action
________ | ______ | __ | ___________________________ | ________ | _______ | ________ | ________ | ________________ | _________ | _________ | _______________ | __________________ | _________________
________ | ______ | __ | ___________________________ | ________ | _______ | ________ | ________ | ________________ | _________ | _________ | _______________ | __________________ | _________________
MATCHED-PERIOD SCORECARD
Eligible matches: ______ Current WPRH: ______ Median WPRH: ______ Trigger minutes: ______
[ ] Context comparable [ ] Safety/food issue escalated [ ] One safe next test selected / no change yet
Hypothesis: _________________________________________________________________________
Owner: __________________ Review date: __________________
Caveats and source note
This is a decision aid, not a risk assessment, food-safety plan, employment model or worker-performance score. Changes to roles, layout or equipment can require a review of the café’s risk assessment. 3 SFBB and the café’s own approved procedures remain in force during busy service. Do not use the map to bypass breaks, cleaning, allergen conversations, temperature checks, supervision or staff rights. Where data identifies a worker, monitoring must be lawful, fair and proportionate, with a clear purpose and appropriate information. 4 Do not claim that the map guarantees faster queues, revenue, staffing reduction, labour savings, compliance or food safety.
Sources reviewed 18 September 2026. The URLs below are the source URLs used in this guide.
References
Checked for this guide
Sources and review date
Primary sources set the compliance and platform guidance. Trade publications are marked as directional signals, not forecasts or guarantees.
- Primarylegislation.gov.uk / The National Archives, The Management of Health and Safety at Work Regulations 1999, Regulation 3: Risk assessment
- PrimaryHealth and Safety Executive, Slips and trips
- PrimaryFood Standards Agency via GOV.UK, Safer food, better business (SFBB)
- PrimaryFood Standards Agency via GOV.UK, Allergen Information for Non-Prepacked Foods Best Practice
- PrimaryInformation Commissioner’s Office, Employment practices and data protection: monitoring workers
