How Much to Tip Delivery Drivers in 2026: Complete Food Delivery Tipping Guide

Delivery tipping is different from restaurant tipping in one crucial way: the money usually shows up before the service is complete. On most food-delivery apps, the tip is part of the offer that helps a driver decide whether your order is worth taking. That means the tip is not just a thank-you after the fact. It is part of the order's economics before the driver ever heads to the restaurant.

That pre-tip mechanism is why low-tip orders wait longer. Drivers are independent enough to decline bad offers, and they are paying their own fuel, maintenance, and time costs. If the app presents an order that looks weak on total payout, experienced drivers often let it pass. This is why the practical delivery question is not only how much is polite. It is also how much gets the order accepted quickly by someone competent.

Platform fees do not solve this problem. The delivery fee, service fee, small-order fee, and regulatory fee are all separate from the tip. Instacart says directly that its service fee is not a tip and does not go to the shopper. Uber's help pages are equally direct that tips go to the courier. So even when checkout already looks expensive, those fees are not a substitute for the worker's tip.

The New York City rules make the mechanics especially clear. After platforms moved tipping later in the checkout flow and average tips fell, the city required restaurant and grocery delivery apps to restore a tipping opportunity before or at the time of order placement. That rule took effect on January 26, 2026, and it requires a suggested option of at least 10%. In other words, the law now treats pre-order tipping as fundamental enough to regulate.

The useful takeaway is simple: delivery tips are part etiquette, part market signal. If you understand that, most of the weirdness around app delivery starts making sense.

The official sources line up on the key pieces. Uber says tips go directly to the courier. NYC's delivery worker FAQ says restaurant and grocery apps must disclose any selected tip to the worker before the worker accepts the offer. The city's January 14, 2026 report explains why that design choice matters.

Small orders are where people get delivery tipping wrong most often. If lunch costs $14, then 15% is only $2.10 and 20% is only $2.80. That may look mathematically tidy, but it does not reflect the actual work: drive to the restaurant, wait for the order, transport it, park, walk it to the door, and reposition for the next job. That is why flat-dollar minimums matter so much more for delivery than for table service.

Once the basket reaches the $20-$40 range, percentage math becomes much more usable. A 15-20% tip now lands in a range that feels like real compensation, especially for a straightforward short-distance order in normal conditions. This is the band where app defaults are most likely to be reasonable, though they still deserve a manual check.

Large restaurant orders need a stronger floor even if you stick with 15-20%. More containers, more drinks, and more handoff friction increase the workload. The same is even more true for groceries. A shopper is not only driving. They are walking aisles, choosing produce, managing substitutions, checking out, and carrying the result.

For very large baskets, a slightly lower percentage can still be fair if the dollar amount is already strong. That is why many practical guides settle around 15-18% on high totals with a high minimum underneath. The point is not to under-tip. The point is to keep the relationship between effort and money reasonable.

Bad weather is the clearest reason to move above your normal baseline. Rain, snow, ice, and extreme heat all make delivery slower, riskier, and less pleasant. If the weather is the reason you ordered in instead of going out, the tip should acknowledge that reality directly.

Distance matters almost as much as weather. A restaurant five or six miles away can absorb a large chunk of a driver's hour once traffic, pickup delay, and the return trip are counted. That is why long-distance delivery should usually get at least a $5 tip even if the order itself was not expensive.

Late-night and holiday deliveries deserve more for a similar reason. The driver pool is thinner, the shift is less convenient, and your order is more likely to require a wait before someone accepts it. The same logic applies when the order is physically annoying: multiple drinks, heavy grocery bags, pet-food sacks, cases of water, or an apartment walk-up all justify extra dollars on top of the base tip.

This is also where delivery differs from restaurant tipping in social feel. Nobody sees you leave a weak delivery tip in real time. The consequence is not public awkwardness. The consequence is slower acceptance, a longer delivery window, or the order being handled by the least selective driver left on the board.

DoorDash follows the core app-delivery pattern: the pre-tip affects how attractive the order looks before a Dasher accepts it. Driver communities are blunt about this. Orders with no tip or a token tip are routinely skipped unless the route is unusually easy. For normal meals, the $4-8 band shows up again and again as the range that feels worth taking.

Uber Eats works in much the same way. SplittyApp's 2026 breakdown reflects the percentages most customers see in the app, and Uber's own help materials say tips go directly to the courier. Uber also says delivery customers can usually edit the tip after the order is complete, which is helpful when service was either much better or much worse than expected.

Instacart deserves a separate rule because the labor is larger. The shopper is doing a picking and substitution job before they ever become a driver. That is why the right benchmark is usually 15-20% with a stronger dollar minimum, especially for heavy orders, apartment buildings, or stores known for long checkout lines.

Grubhub and smaller delivery apps are not structurally different enough to justify a different etiquette standard. The same basic rule works almost everywhere: use a meaningful pre-tip, protect the worker on small orders with a floor, and adjust upward for weather, distance, or complexity.

For official platform mechanics, Uber's customer help, Uber's courier help, and Instacart's tip help are the best primary references. For the NYC-specific legal overlay, rely on the city's own guidance first.

The strongest argument for pre-tipping is not moral. It is operational. Drivers sort through offers quickly, and the ones that look profitable get accepted first. If your tip is weak, your order becomes something the market has to rescue rather than something a driver wants immediately.

The table below is not a formal platform statistic. It is a practical summary of recurring patterns from driver discussions, especially in Reddit's DoorDash communities. Treat it as field guidance, not a guaranteed probability model. Even so, it captures the main reality accurately: there is a steep difference between a no-tip order and a $5-8 order.

Most platforms let customers increase a tip after a great delivery, and some let them reduce or edit it after a bad one. Use that carefully. The fair reason to cut a tip is a real delivery problem such as careless handling, ignoring instructions, or a completely wrong drop-off. A soggy burger or a missing side is usually a restaurant problem, not a driver problem.

Tip baiting is the failure mode every serious delivery guide should name directly. That is the practice of pre-tipping high to get accepted quickly, then reducing the tip after the order arrives. Drivers hate it for obvious reasons, and it damages the trust that makes pre-tipping work at all.

Delivery drivers and shoppers live inside a cost structure most customers never see. They bring the car, the fuel, the insurance exposure, the maintenance, the parking risk, and the dead time between offers. On grocery apps, they also bring the labor of navigating the store and solving inventory problems in real time.

That is why the tip matters more here than it does in many other service categories. The platform fee stack can be large, but the worker does not automatically benefit from those fees. The tip is one of the few checkout lines that is transparently meant for the person doing the delivery.

The New York City reporting around delivery tipping makes another point clear: when apps push the tip later or make it less visible, both tip size and worker earnings can fall. That means the pre-tip is not some sentimental leftover from old tipping culture. It is part of the system that decides whether delivery labor remains viable order by order.

If you want one economic rule to remember, it is this: delivery pay is driven by total payout, not by your basket total alone. That is why small orders need a stronger minimum, and that is why a seemingly small difference between a $2 tip and a $5 tip changes the worker's decision so much.

The cleanest official proof is still the simplest: Instacart says its service fee is not a tip, and Uber says the courier gets 100% of the tip. That is enough to tell you which checkout line is actually paying the worker.