Last Updated: August 8, 2026 | The platform’s order assignment interface, rewards, and order acceptance rules are subject to change based on your account, city, and time. Please refer to the information currently displayed in the delivery app and official announcements for the most up-to-date details.
The most common mistake during peak hours isn’t necessarily accepting the wrong order, but accepting every order that comes along. Although orders come in thick and fast during lunch and dinner hours, restaurant preparation times, traffic, parking, and elevator wait times also slow down. If you only look at the pay shown on the screen without factoring in the total time the trip takes, you’ll likely end up working harder than usual, yet your actual hourly earnings won’t be any higher. A more reliable approach is to treat each order as a small decision involving “time, distance, and risk,” rather than simply focusing on whether the pay is high.
This article outlines a peak-hour order-acceptance process that you can test for yourself. The focus isn’t on teaching you to aim for a fixed price per kilometer, nor is it a guarantee that a certain driving strategy will always be profitable. Instead, it’s designed to help you quickly assess—in just a few seconds—how much time a ride might take, where it will take you, and whether you’ll be able to accept another ride smoothly after completing it. For new drivers, starting with simple records to establish your own baseline is usually more practical than blindly copying someone else’s standards.
Estimate the total time first; don't just look at the kilometers displayed on the screen.
When evaluating an order, you can break it down into five stages: traveling to the restaurant, waiting for the food inside the restaurant, packing and picking up the order, delivering it to the customer, and returning to the order-acceptance area after completion. The distance displayed by the platform usually reflects only a portion of this process and does not fully account for traffic lights, locating the correct address, community access checks, going up stairs, and the return trip. During peak hours, if the restaurant is swamped with orders, even an order that initially seems very close can end up taking a significant amount of time just to wait.
The simplest way to estimate is to first ask yourself three questions: How long will it take me to get to the store? How long is the current wait time at this store? Are there any follow-up orders at the delivery location after I drop off this one? If you estimate the entire trip will take 25 minutes, divide the pay by 25, then multiply by 60 to get a rough “hourly pay for the duration of the trip.” This isn’t your actual hourly wage, since it doesn’t account for gas, maintenance, supplies, or downtime, but it can help you compare which of two orders is more efficient.
For example, suppose two orders both show a fare of 100 yuan: one is estimated to take 20 minutes and is in a high-demand area, while the other is estimated to take 35 minutes and is in a remote residential area. The value of these two orders is not the same. The estimated hourly pay for the former is 300 yuan, while that for the latter is approximately 171 yuan—and the latter may also require a return trip. What really matters is how long it takes to complete an entire cycle, rather than just looking at the order amount.
Include store wait times in your personal database
For orders over the same distance, the difference often comes down to the restaurant. Popular restaurants may be handling dine-in, takeout, and multi-platform orders simultaneously during lunch and dinner hours; some restaurants prepare meals on time, while others wait until the delivery driver arrives before starting to prepare them. Instead of complaining based on vague impressions, use your phone’s notes app to record the date, time, arrival time at the restaurant, pickup time, and the reason for any issues. After collecting data for one to two weeks, you can then make an informed judgment.
When keeping records, don’t be quick to assume that a single delay represents a consistent pattern. It could be due to a temporary staff shortage, special meal requests, system delays, or rain that day. Only add a store to your list of high-risk peak hours if it experiences long wait times multiple times during similar time periods. If you’ve already accepted an order and encounter a long wait, first follow the app’s procedures, save a screenshot and record the time, and then decide how to proceed; you can also use this site’sGuide to Recording and Handling Customer Wait Times...to avoid making the situation even more tense by letting emotions get in the way of communication.
A restaurant database doesn’t have to be complicated. For each restaurant, it’s enough to record just three items: “average wait time,” “parking availability,” and “peak-hour crowding.” Once you’re familiar with the area, you’ll be able to quickly adjust the system’s estimated time just by seeing the restaurant’s name—which is usually more accurate than setting a citywide threshold per kilometer.
When combining orders, first consider the workflow, then how much the commission will increase.
Whether combining orders is worth it depends not on delivering two orders at once, but on how much extra time the second order adds. Combining orders can only improve efficiency if the two pickup locations are close to each other, the delivery routes are in the same direction, and the meals can be prepared within a reasonable timeframe; if the route requires crossing major thoroughfares, backtracking, or if one of the restaurants is particularly slow to prepare orders, the additional compensation may not be enough to offset the delay.
When you see a combined order, check the following four things in order: whether the new pickup location is on your route; the risk of delays at both locations; whether the delivery order makes sense; and whether the first customer’s order will sit in the car for too long. You should also take the space in your insulated bag into account—if large drinks, soups, and pizzas are crammed together, the risk of spills or food damage increases significantly. When space is limited, turning down one order is often more time-efficient than having to pay compensation or handle customer complaints afterward.
The information displayed on the platform, the ability to decline additional orders, and the procedures following a cancellation are all subject to change. Do not treat any individual’s order acceptance or cancellation rate thresholds found online as a general rule, and do not assume that skipping a particular order will necessarily affect the next assignment. You can rely on the information currently displayed in the app, your partnership agreement, and official guidelines; it’s best not to base major decisions on unverified rumors about the algorithm.
The delivery location determines the opportunity cost of the next order
Long-distance rides during peak hours aren’t necessarily bad; the issue lies in the driver’s location after dropping off the passenger. If the destination is near a commercial district, a train station, or a popular dinner spot, the driver may be able to pick up the next ride quickly after completing the trip. However, if the destination is in a mountainous area, an industrial zone, or on the outskirts of a large residential area—where orders are scarce—the time spent returning must be factored into the cost. This revenue-free travel time is often the reason why reported earnings look good on paper, but actual hourly earnings end up being lower.
In familiar areas, you can identify three types of locations: spots where orders come in consistently, spots prone to traffic jams, and spots where it’s difficult to return after delivery. When you receive a cross-district order, first determine which category the destination falls into, then assess whether the pay is sufficient to cover the return trip. Don’t park illegally or make dangerous U-turns just to stake out the so-called “hot spots”; areas where you can find legal places to wait and enter and exit safely usually yield better long-term efficiency.
Weather can also affect location value. During rain, bridges, underpasses, and shopping mall driveways may become congested; in the summer, be sure to schedule time for hydration, cooling off, and rest during midday. If fatigue is already affecting your judgment, it’s not worth the risk—no matter how high the peak-hour bonus. The losses resulting from accidents and injuries far exceed the loss of a single ride; you may want to review your insurance coverage in advance.A Guide to Insurance for Food Delivery Drivers...and your own insurance policy and the platform's supporting documents shall prevail.
Use a 7-day tracking exercise to identify your threshold for accepting orders
The threshold for accepting rides shouldn’t be based solely on “X yuan per kilometer.” We recommend keeping track of the following for seven consecutive days: hours online, number of completed rides, fare earnings, bonuses, tips, total mileage, fuel or charging costs, parking fees, and any obvious idle time. At the end of each day, calculate two figures: total earnings divided by total online hours, and the net hourly rate after deducting directly identifiable costs. This is the only way to determine whether a particular order-taking strategy actually improves your earnings.
Next, review the low-efficiency orders and briefly note the reasons, such as “waited more than 15 minutes for the meal,” “delivered to a remote area,” “took too long to find the address,” “had to turn back after combining orders,” or “rush-hour traffic.” If the same reason recurs frequently, incorporate it into the decision-making criteria for the following week. For example, instead of refusing all long-distance orders outright, raise the acceptance threshold for orders that “are delivered to a remote area and require an empty return trip”; instead of rejecting all combined orders, avoid combinations involving two high-risk stores with long wait times.
Under the current law governing food delivery services, the base pay for each order must be calculated proportionally based on the duration of the delivery service, must not be less than 1.25 times the minimum hourly wage, and must not be less than 45 yuan per order; for details on the actual calculation and applicable conditions, please refer toExplanation of Compensation on the Ministry of Labor’s Special Section for the Protection of Food Delivery Riders’ Rights. The statutory minimum is not the same as your personal net income goal. You still need to keep track of fuel costs, vehicle depreciation, maintenance, and waiting time yourself, so you don’t mistake revenue for actual profit.
The bottom line for peak efficiency is safety and traceability
No matter how refined your order-acceptance strategy is, you must never run red lights, ride against traffic, use your phone while riding, or park illegally to save time. During peak hours, when there’s a lot of information to process and pressure is high, you should secure your phone, adjust the navigation volume, and stop in a safe location before confirming a new order. If disputes arise regarding meal delivery, house numbers, customer instructions, or waiting at a restaurant, saving the necessary screenshots, communication records, and timestamps will be more helpful than relying on memory to explain the situation afterward.
If you encounter issues such as incorrect deliveries, damaged food, customers who cannot be contacted, or billing disputes, first follow the platform’s customer service and appeal procedures. Do not make compensation promises on your own that exceed your capabilities, and avoid posting customers’ personal information on public social media platforms. This site’sGuide to Resolving Food Delivery DisputesI've organized the records in chronological order; if there are any issues related to suspensions, contracts, or compensation, they can also be tracked.Food Delivery News and Policy Updates...Subject to government announcements and official notices from the platform.
During peak hours, what really makes the difference isn’t how fast you tap, but whether your time estimates are close to reality. First, look at the total trip time; then examine the pickup location and destination; next, assess the increased risk of combining orders; and finally, adjust your threshold based on your own seven-day records. When you can clearly explain why you accepted a particular order and whether the outcome met your expectations, you’ve moved beyond relying on gut instinct and toward a sustainable, adaptable approach to accepting orders.
This article is a compilation of general experiences and information and does not constitute a guarantee of platform earnings, legal advice, or insurance advice. The rules, order-assignment screens, and rewards for each platform are subject to change at any time; please refer to the app, partnership agreements, announcements from regulatory authorities, and your own actual records for details.
