How to Handle Busy Periods

A busy period becomes damaging when orders enter faster than the kitchen, packing area or delivery operation can complete them. The solution is not simply to work faster. It is to control demand, protect the bottleneck and give customers a service time the business can realistically meet.

Measure the real bottleneck

Start with recent busy shifts. For each 15- or 30-minute period, record:

• orders received by channel;
• items or preparation workload, not only order count;
• promised collection or delivery time;
• actual ready and delivery time;
• the station where work accumulated;
• driver availability and round-trip time;
• refunds, remakes, missing items and complaints;
• when staff paused orders or extended lead times.

Order count alone can be misleading. Ten simple collection orders may require less capacity than five large mixed orders with delivery, meal deals and several modifiers.

Set a service promise the kitchen can control

Choose a target that staff can maintain on a typical busy shift, then define the point at which the promise changes. For example, the business may move from normal lead time to extended lead time, then to restricted menu or paused orders.

Do not copy a generic threshold such as “pause at 75% capacity”. Capacity is not measured consistently between kitchens. Establish the trigger from actual queue length, workload at the bottleneck station, overdue orders and delivery resources.

Operating state	Possible trigger	Action
Normal	Orders completing within the promised range	Use standard slots and menu
Busy	Queue building but service promise still achievable	Extend lead time, reduce slot availability and prepare to restrict slow items
Constrained	Bottleneck at limit or orders approaching late status	Restrict menu, cap delivery area or pause selected channels
Recovery	Queue falling and overdue orders cleared	Reopen gradually rather than restoring every channel at once

Design time slots around workload

A time slot is a capacity control, not merely a choice shown at checkout. If every customer can select the same half-hour period, the slot does not protect the kitchen.

Set limits using representative order workload. Consider:

• the number of fryer, grill, oven or wok positions;
• packing capacity;
• average and unusually large order sizes;
• the mix of collection and delivery;
• walk-in and telephone demand that is not visible online;
• handover space and collection congestion;
• driver round-trip capacity.

Where the system allows it, stagger collection promises instead of directing every customer to arrive at the start of a broad window. For delivery, the slot must account for both kitchen completion and available drivers.

Use pre-orders carefully

Pre-orders can improve planning, but they do not create extra capacity. A large group of orders scheduled for 19:00 can overload the same stations as orders placed at 18:45.

Control pre-orders by:

• limiting the number or workload accepted into each slot;
• reserving capacity for walk-in, telephone and immediate orders;
• checking stock and staffing before opening long advance windows;
• separating preparation start time from customer collection time;
• making large-order rules clear;
• reviewing pre-orders before service begins.

Throttle before the queue becomes unmanageable

Throttling means reducing the rate or complexity of new orders. It protects the service already promised to customers.

Options include:

• extending the next available collection or delivery time;
• reducing the number of available slots;
• pausing one channel while another remains open;
• temporarily removing items that overload the bottleneck station;
• reducing the delivery area or number of simultaneous deliveries;
• switching delivery orders to collection only where the system and customer communication support it;
• temporarily stopping new orders.

The trigger and authorised person should be agreed before service. Staff should not have to wait for an unavailable owner while late orders continue to accumulate.

Create a peak-period menu

A shorter menu can protect throughput when it is designed in advance. Review which items:

• occupy the bottleneck equipment for too long;
• require extensive customisation;
• are difficult to pack or travel poorly;
• depend on an ingredient that often runs out;
• create a disproportionate number of mistakes;
• have low demand and interrupt the main production flow.

Do not remove items unpredictably without checking meal deals, modifiers and promotions. Test the restricted menu before using it during a live peak.

Keep wait times consistent across channels

Customers are more likely to accept a long wait when it is stated before payment. They are less likely to accept a short promise that is repeatedly missed.

Assign responsibility for updating:

• website and app lead times;
• marketplace preparation settings where available;
• telephone estimates;
• collection instructions;
• delivery status messages;
• in-store notices.

Marketplace controls differ by provider. Verify what can be changed manually, what the platform calculates and how pauses or delays affect customers and drivers.

If an accepted order will be late, contact the customer before the promised time where practical. Give a realistic update and an available remedy rather than a vague assurance that the order is “on the way”.

Allocate staff to the constraint

More staff do not automatically produce more output. Place labour where work is accumulating.

• Restock and stage packaging before the peak.
• Move cleaning and non-urgent administration outside the critical window.
• Separate order monitoring, customer handover and kitchen production where staffing allows.
• Assign one person to watch digital orders, sold-out items and status exceptions.
• Use an expediter or packer when missing items and handover errors become the constraint.
• Cross-train staff for the stations most likely to become overloaded.

A written pre-peak checklist is more reliable than relying on experienced staff to remember every preparation task.

Calculate delivery capacity separately

Kitchen capacity and delivery capacity are different. A meal can be ready on time and still arrive late because no driver is available.

Estimate driver capacity from actual round-trip time, including waiting, parking, handover and return. For planning purposes:

Approximate deliveries per driver per hour = 60 divided by average round-trip minutes

This is only a starting estimate. It must be reduced for gaps between runs, grouped deliveries, traffic variation and difficult addresses. Use recent delivery records rather than an ideal journey time.

When delivery becomes the constraint, options include extending delivery promises, narrowing zones, increasing minimum orders where commercially sensible, limiting simultaneous bookings, adding suitable driver cover or promoting collection. Do not continue accepting delivery orders solely because the kitchen can still prepare them.

Prepare failure procedures

Busy periods magnify small failures. Staff need a simple response for:

• lost internet connection;
• printer or KDS failure;
• duplicate or missing orders;
• payment accepted but no kitchen ticket;
• an unavailable item after acceptance;
• driver shortage;
• an unsafe or incomplete delivery address;
• a customer who cannot be contacted.

Keep the fallback visible and short. It should state who decides, how the customer is contacted, how the order is recorded and how payment or refund action is completed later.

Review the shift while the evidence is fresh

After a peak, hold a brief review. Focus on observable facts:

• when the queue began to grow;
• which station or resource became the constraint;
• whether lead times were changed early enough;
• which items caused delay or errors;
• whether customers received accurate updates;
• how many interventions and refunds were required;
• which change should be tested on the next comparable shift.

Change one or two controls at a time so the result can be evaluated. The aim is not to accept the largest possible number of orders. It is to complete a profitable level of demand accurately and within an honest service promise.

Related guides

• Online Ordering — the main guide for the wider topic.
• Order Handling and Operations — the broader guide that frames this implementation.
• Pre-Ordering and Advance Orders — a closely related operational control to review alongside this page.
• Order Accuracy and Quality Control — a closely related operational control to review alongside this page.