When your restaurant is short-staffed, restrict the fulfillment mode that puts the most pressure on the scarce part of the shift—not automatically delivery or pickup. If the kitchen is the bottleneck, limiting either mode may help. If the counter is overwhelmed but couriers are available, delivery may take less staff handoff time. If dispatch is unreliable or expensive, pickup may be the better choice.
Make the decision for a specific service window, not forever. Estimate how many orders the team can prepare and hand off, compare what each mode contributes after its variable costs, and leave room for problems. Then tell guests what is available before they order.
Estimate the labor each mode needs
Start with the people and time actually available. The shift lead can identify who is absent, who can cover each station, and how much time remains after dine-in service and essential closing work. The kitchen lead can estimate preparation and packing time. The person handling the counter or dispatch can estimate handoff work and identify recurring delays.
Count the work an order creates across the whole team:
- Kitchen: reading the ticket, preparing food, packing it, and checking accuracy.
- Pickup handoff: staging bags, answering arrival questions, locating orders, and resolving missing items.
- Delivery handoff: arranging dispatch where needed, staging bags, checking the courier's order, and handling a failed pickup.
- Exceptions: remakes, late arrivals, unavailable items, and guest or courier questions.
Some work is shared by both modes. A pickup order does not avoid cooking and packing, and a courier does not eliminate the restaurant's responsibility to prepare and hand over the right bag. Measure or estimate staff minutes separately from elapsed time: a courier's trip may take 20 minutes without occupying a restaurant employee for all 20.
Use a recent comparable shift if you have one. Otherwise, ask the people doing the work for a conservative estimate and update it after service. Check whether large orders or particular menu items need more time than an average order. An average is useful for setting a starting limit, but it can hide a rush of complicated tickets.
Also identify the delivery resources outside the restaurant. Who arranges the courier? How many pickups can realistically be covered during the window? What happens when a courier is late or unavailable? Get the dispatch fee and any cancellation or exception charges in writing. If you use your own drivers, account for their available time and costs rather than treating them as an unlimited resource.
Compare capacity before choosing a restriction
Capacity is the smallest limit among the resources an order needs. For pickup, that may be kitchen or counter time. For delivery, it may be kitchen time, dispatch work, or available couriers. Subtract time already committed to dine-in and other work before setting an online-order limit.
The following example is hypothetical. It covers one busy hour after other commitments. The team has 60 kitchen minutes and 30 handoff minutes available. Each order uses six kitchen minutes. A pickup order needs three handoff minutes; a delivery order needs one. Assume couriers can handle at most six deliveries in the hour. These are planning assumptions, not a promise of performance.
| Option | Order capacity | Resource check | Decision use |
|---|---|---|---|
| Pickup only | 10 orders | 60 kitchen minutes; 30 handoff minutes | Uses all available kitchen and handoff time. |
| Delivery only | 6 orders | 36 kitchen minutes; 6 handoff minutes; six courier slots | Courier availability limits orders before the kitchen does. |
| Six pickup, four delivery | 10 orders | 60 kitchen minutes; 22 handoff minutes; four courier slots | Preserves both modes but leaves no kitchen buffer. |
The table shows why “delivery takes less counter time” is not enough to choose delivery alone. Under these assumptions, restricting pickup to favor delivery would reduce the maximum accepted orders from 10 to six because courier capacity is tighter. A mix can reach 10, but it leaves no kitchen time for a remake or an unexpectedly complex order.
A practical starting decision would be to keep both modes available with a combined limit below 10, subject to the delivery-slot limit. For example, five pickup and three delivery orders use 48 kitchen minutes, 18 handoff minutes, and three courier slots. That leaves 12 kitchen minutes and 12 handoff minutes for variation. It is a starting buffer, not a guarantee that eight orders will run smoothly.
If the counter becomes the real bottleneck, reduce pickup slots or shift some capacity to delivery only if courier coverage and kitchen capacity allow it. If couriers become unreliable, reduce delivery slots and favor pickup only if the counter has room. If the kitchen is behind, reduce the combined order limit regardless of mode. A longer quoted wait can spread demand, but it does not create labor.
Before relying on any limit, ask your ordering provider how it handles capacity: Can you set separate pickup and delivery hours or limits? Can you control the total across both modes? If the setup cannot enforce a combined limit, the shift lead needs another reliable way to monitor accepted orders and stop taking more before the team is overloaded.
Tell guests what changes and when
Set a clear start and end time for the temporary arrangement. Update the ordering screen and any other places where guests may begin an order. Your restaurant website can help guests find the current options before they reach checkout. Ask your provider where changes must be made and whether they apply immediately to orders already in progress.
Use a message that names the available service, the affected window, and the next step. For example:
Tonight from 5–7 p.m., pickup and delivery slots are limited while we operate with a smaller team. Please choose an available time at checkout. If no time appears, try a later window.
If one mode is unavailable, say which one and for how long. Do not advertise a pickup alternative unless it is actually open and has capacity. Give the staff answering calls the same wording, including what to say when all slots are full. Avoid promising a specific reopening time unless someone is responsible for confirming it.
Check the guest journey before service. In a simulated practice scenario, have a staff member walk through selecting each mode, viewing available times, and reaching the point where the temporary message should appear. Stop before any live payment, guest message, or courier dispatch. If your system offers a genuine test environment, use it; otherwise, ask the provider how to check safely without creating a real order.
A simulation can reveal confusing wording or a setting that does not behave as expected. It cannot prove the team will keep up during a rush. A live service pilot is different: apply the agreed limits to real orders during a defined window, brief the team, and monitor the shift. Only genuine customer transactions and normal fulfillment should occur during that pilot. Give the shift lead authority to tighten the limits if the queue grows.
Compare contribution as well as throughput
More accepted orders are not automatically better if their extra costs consume the sale. Compare each mode using the revenue and costs that change with that order. Include food, packaging, processing, incremental labor, and any dispatch expense. Count a delivery charge collected from the guest as revenue, while recognizing that it is also a cost to that guest.
Here is a separate hypothetical per-order comparison. Assume a $24 food sale, $10 in food and packaging, and $1 in payment processing for either mode. Pickup adds $3 of incremental labor. Delivery adds $1 of incremental labor and a $6 dispatch charge; the guest pays a $2 delivery charge.
- Pickup contribution: $24 − $10 − $1 − $3 = $10.
- Delivery contribution: $24 + $2 − $10 − $1 − $1 − $6 = $8.
These figures include the stated variable costs. They exclude fixed costs such as rent and subscriptions, as well as exception costs such as remakes, failed dispatches, or refunds. They are not profit figures or quotes. If labor would be paid for the shift regardless of order volume, do not treat all scheduled wages as incremental without checking the assumption. Conversely, include extra paid time if accepting more orders causes it.
In this example, pickup contributes $2 more per order under the stated assumptions. That alone does not settle the staffing decision: pickup may use more of a scarce handoff resource, while delivery may have fewer available courier slots. Compare contribution with the capacity calculation for the same service window. Do not combine figures from the two examples as though they describe one restaurant.
For an actual quote, separate the software subscription from per-order processing, dispatch, and other charges. As of November 2026, Fleksa's plans list Fix at $199 per month per location with 0% Fleksa commission on direct online orders. Processing, delivery or dispatch, hardware and add-ons, and applicable sales tax are extra. A commission-free direct channel does not mean a free courier. Confirm current terms and which charges your restaurant and guests each pay.
For a broader comparison of the costs that change by channel, see the direct-order contribution guide. Keep that calculation separate from the immediate question of how many orders today's team can fulfill.
Review the shift and adjust the next limit
At the end of the window, ask the shift lead for a short record: orders accepted by mode, quoted versus actual ready times, late or failed handoffs, remakes, refunds, and times when a station fell behind. Note whether staff had to stay late. Compare the record with the estimates used to set the limit.
Look for the constraint, not just the busiest person. A pile of finished bags waiting for couriers points to a different adjustment than tickets waiting to be cooked. A crowded counter may call for fewer pickup slots or a clearer staging process. If both modes were late because the kitchen was full, lower the combined limit or simplify the available menu where your setup allows it.
Review the customer experience too. Did guests see the restriction before trying to order? Were quoted times credible? Did the team know what to say when capacity ran out? A limit that protects the kitchen but sends guests through a confusing checkout still needs work.
For your next shift, write down the scarce resource, available minutes, mode-specific work, courier limit, proposed order cap, and the person authorized to change it. Use that checklist to evaluate your current setup, then ask Fleksa about restaurant online ordering, the controls you need, and verified availability in your US location.
Frequently asked questions
Should I turn off delivery first when staff are short?
Not automatically. Delivery may require less counter time, but it still uses kitchen labor and depends on courier capacity. Restrict the mode that strains the resource you lack, while accounting for its contribution and guest demand.
How many online orders should I accept?
Estimate available kitchen and handoff time after other commitments, then divide by the time each order requires. Apply any courier limit and leave a buffer for variation. Review actual service results before using the same cap again.
Can a longer pickup or delivery wait replace an order limit?
A longer quote can spread orders across available times, but it does not add staff or courier capacity. Check that the promised times match what the team can fulfill, and restrict slots when the workload still exceeds capacity.
