Operations

Downloaded 28 September 2026

Delivery Operations SOP

Delivery doesn't ruin kitchens. Unmanaged delivery does. Here's how to run it without burning down your dine-in service.

What This Is

Delivery operations SOP covers how you manage the full delivery flow: aggregator platform management, packaging, ticket time targets, kitchen sequencing, and the decisions around when to throttle or cut off delivery to protect dine-in quality. It applies equally to cloud kitchens and to dine-in restaurants with a delivery channel.

What Most Operations Get Wrong

Four specific ways delivery kills kitchens, in descending order of how much damage they do:

1. Delivery and dine-in compete for the same kitchen capacity at the same time. You have one grill. A delivery ticket and a dine-in ticket arrive simultaneously. The grill cook can only run one. If there’s no priority rule, they make a random decision every time, and dine-in guests wait while the delivery order cooks, or the delivery rider waits while the dine-in order goes first. Neither is managed. Both are suffered.

2. Packaging is treated as a detail. Biryani in a leaking container, a burger that arrived in a box two sizes too large and shifted into a structural collapse, a salad that arrived as a liquid, these are not packaging accidents. They are decisions not made. Every item on your delivery menu needs a packaging decision as deliberate as its recipe.

3. Aggregator menus are not maintained. Items that sell out at 8pm are still live on the platform at 9pm. The kitchen starts rejecting orders. Rejection rate goes up. Zomato and Swiggy penalise rejection rate, your visibility drops. More rejections follow.

4. There is no delivery ticket time target. On dine-in, you’d never accept unlimited ticket times. On delivery, most operations have no target, the kitchen just works through whatever is in the queue. Without a target, there is no standard. Without a standard, there is no improvement.

The Process

Platform Setup and Menu Management

Maintain a separate delivery menu. Not every item on your dine-in menu should be available on delivery. Ask one question for each item: does this travel well and arrive at the same quality as when it leaves the kitchen? If the answer is no, a hand-pressed dosa, a souffle, a salad that wilts in 20 minutes, do not list it. A shorter delivery menu with high consistency beats a full menu where half the items disappoint at the door.

Build platform management into someone’s daily job. Specifically: checking and updating item availability before each service window, monitoring rejection rate (should be under 5%), responding to platform support tickets within 24 hours. This person does not need to be senior, they need to be consistent.

Know your aggregator economics. Swiggy and Zomato take 15-25% commission plus applicable taxes. Add your packaging cost (typically ₹15-40 per order depending on items). Add the cost of a delivery-only team member if you have one. Now calculate what percentage of your delivery revenue actually reaches you as contribution. If it’s under 55%, your delivery pricing needs review.

Ticket Time Targets

Set a kitchen-to-handoff ticket time target for delivery. A reasonable starting target for most operations: 18 minutes from order confirmation to order ready for pickup. This includes all prep and packaging. If your average is above 25 minutes, you are accumulating cancellations and bad ratings.

Measure it daily for the first month using your aggregator dashboard. Both Swiggy and Zomato provide preparation time data in their partner dashboards. This data is your performance metric.

When to adjust the target: During peak windows (Friday and Saturday 7-10pm), extend the target to 22 minutes and communicate updated preparation time to the platform, proactively, before orders stack up. The platform’s estimated delivery time to the customer is based on your stated prep time. Honesty here maintains ratings; optimism here destroys them.

Kitchen Sequencing: Delivery vs. Dine-In

The non-negotiable rule: dine-in gets kitchen priority during active service. Delivery orders are batched and sequenced in between dine-in tickets, not alongside them.

Practical implementation:

The cutoff: If your kitchen is at 90%+ capacity during service, activate your delivery pause. Both Swiggy and Zomato have a “pause orders” function. Use it. A 30-minute pause during a slammed Saturday dinner saves your kitchen from a 45-minute spiral of backed-up orders, late deliveries, and 1-star ratings. Train your floor manager to make this call, do not leave it to the kitchen team to flag upward.

Packaging Standards

For every delivery item, define:

Build a packaging spec sheet. Photograph the correct pack for each item. Show it to the kitchen team in training. Check it during service spot-inspections, not just when a complaint arrives.

The 4 Delivery Metrics to Track Daily

  1. Average ticket time (kitchen confirmation to ready for pickup)
  2. Rejection rate (cancelled by kitchen / total orders)
  3. Average rating on each platform
  4. Order volume by hour , to identify your peak window and staff accordingly

Review these weekly. Any metric that has deteriorated week-on-week needs a specific cause identified and a specific action to address it.

Sector Notes

Independent Cafe

Delivery is probably not your core channel, and it shouldn’t be. Cafe items that travel poorly include most coffee drinks, any pastry with a cream or fresh element, and open sandwiches. If you are going to offer delivery from a cafe, constrain your delivery menu to sealed beverages (cold brew, bottled items), packaged baked goods, and any food that holds well. A chai latte delivered 30 minutes late is a brand experience problem.

Restaurant (Full Service)

The risk for full-service restaurants adding delivery is real. Your kitchen is designed and staffed for a specific cover count and ticket timing. Adding delivery volume without adding kitchen capacity or rationalising your delivery menu is adding load to a system that was not designed for it. Run a delivery audit quarterly: measure whether your dine-in ticket times degraded during periods of high delivery volume. If they did, your delivery channel is cannibalising your dine-in quality.

Hotel F&B

Hotel delivery is typically in-room dining, a controlled, internal channel. Delivery to external addresses via aggregator platforms is unusual for hotels and generally inadvisable for full-service outlets. However, if a hotel has a standalone F&B brand that delivers, the economics and standards above apply. For in-room dining, the ticket time target is typically longer (30-35 minutes) but the quality standard is higher, packaging presentation and tray setup matter as much as the food.

Cloud Kitchen / Delivery

This entire SOP was written for you. A few specific additions: your brand is your packaging and your delivery experience, invest here disproportionately relative to dine-in operators, because it is the only physical brand touchpoint your customer has. Build a QC checkpoint before every order exits the kitchen: one person who is not the cook checks every order against the ticket before packaging is sealed. This adds 45 seconds per order and eliminates missing-item complaints.

QSR / Fast Casual

Delivery from QSR is competitive because the product often travels well (burgers, wraps, bowls) and the price point is right for delivery. Your risk is in the aggregator’s handling time, by the time the order reaches the customer, 20 minutes of travel time has affected your fried items. Design your delivery menu around items that hold for 25-30 minutes without quality loss, even if that means holding back your most popular dine-in item from the delivery channel. A good delivery version of your brand beats a bad delivery version of your full menu.

How You’ll Know It’s Working

Your average delivery rating on Zomato and Swiggy will be above 4.2 and stable, not oscillating based on peak day chaos.

Your dine-in ticket times will not increase on days when delivery volume is high, because the two channels are being managed separately, not competing.

Your rejection rate will be under 3% consistently, because your platform menu is up to date and your capacity decisions are being made proactively.