Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Something is going to need to give in food delivery because nobody is making a profit on it.

Restaurant loses money on the percentage.

Driver loses money on the vehicle depreciation much of the time.

The delivery company loses money on the delivery.

I just don’t think a burger can be profitably delivered 15 minutes away for 1.99.



"Restaurant loses money on the percentage."

Do they? For delivery, FOH is out of the equation, so there is a very big cost of running a restaurant that is not impacted.

"The delivery company loses money on the delivery."

They're a software/customer support service. At scale there has to be an inflection point.

"I just don’t think a burger can be profitably delivered 15 minutes away for 1.99."

It's way more than that... 1) the upfront cost the person pays (these services are sneaky about how this is calculated, often times the menu prices are adjusted for delivery + the hard fee) 2) the amount the restaurant pays 3) tip for the driver. The end user will end up paying $5 for say a $10 meal, and the restaurant might pay $2.

I guess the question is - does a person want to spend 30 minutes driving + gas vs. pay $5 for the convenience of delivery for a $10 meal? I think quite a few people will pick the latter.


Restaurants aren't software though. The FOH cost doesn't disappear like terminated EC2 instances. The sometimes-lavish furnishings and rent on a pricy, high-profile location with a lot of foot traffic remain, even after firing (human) servers/waiters and other FOH staff.

Ghost restaurants, which are restaurants run out of commercial kitchens away from downtown areas, and with no physical presence to speak of, using the Internet to drive orders, are trying out running FOH-less operations, but for an existing restaurant that's a bit more difficult a move to make. (Since a restaurant with ~no FOH staff has no human servers/waiters, maybe we could call that model serverless.)


Restaurants and delivery contractors should figure out very quickly if they come out positive on this. Maybe the restaurants won't figure out how much was cannibalization, but they'll know if it's a net loss. If you're losing money with, say Grubhub, you stop working with Grubhub.

The only question is if delivery services can find scale and a price point where they're profitable.


What do the economics look like for the old Pizza or Chinese delivery places in a pre-app world?


I did delivery for a slightly upscale pizza restaurant about ten years ago. For lower tier chain restaurants I think delivery driving was a scam in which the restaurant relied on the employee being unable to calculate that costs of gas + vehicle depreciation made the effective wage of driving delivery very close to zero.

I think the old world was restaurants squeezing delivery drivers. Now it's the app squeezing both the restaurant and the delivery driver.


Perhaps this is a bit dated, but one of my relatives once did delivery for a Chinese restaurant. He worked solely for tips. So in that case, the entire delivery was paid for by the customer.

Not so sure about pizza, but the tip issue probably comes into play there again. The customer covered the delivery cost and a 15% tip is more than most delivery fees in my area even on small orders.


Restaurant: Keeps ~85-90% of order

Delivery driver: Paid a portion by restaurant, paid a portion by tips

Telephone: $60 / month utility bill

----

Today:

Restaurant: Keeps 60-70% of order

Delivery driver: Paid a portion by Uber, paid a portion by tips

Uber: Paid 30% + delivery fees, service fees

See the problem?


To be fair, they often also pushed depreciation onto their (young, naive) drivers.

But other than that, there is essentially no marginal cost to the delivery infrastructure, you pay part of the driver cost directly (via tip), they at most take a cut to go to drivers (minimum) wages, but that's a fixed cost not % of checks.


Lots of those roles paid/pay under the table. At GrubHub's scale, tax evasion doesn't work.


Ancedotally (I don't have a ton of concrete details here):

1. From what I understand restaurants doing delivery largely were breaking even on it considering the costs - there was no third party looking to take 20-30% of the gross.

2. Again anecdotally, I suspect the old school pizza/chinese delivery places are more efficient in terms of deliveries per driver per hour. What I see with a lot of UberEats/Grubhub etc is drivers waiting a lot at restaurants, bouncing around a bunch to pick up one or 2 orders each. Vs the old-school pizza model of a driver being able to pick up a bunch of orders at once.


> the old school pizza/chinese delivery places are more efficient in terms of deliveries per driver per hour

Precisely this! It's amazingly inefficient to dedicate an entire person (and their 1000kg car) for the delivery of a single 2kg takeaway meal.

Uber tries to "solve" the problem by giving drivers two orders at the same time, but their implementation is terrible:

1. Sometimes the restaurant doesn't have both orders ready at the same time, so one of the orders is growing cold and soggy.

2. Uber's dispatch system has such a strong preference for batching deliveries together that a driver will often get assigned two orders that are going in opposite directions from the restaurant!

3. The high-volume restaurants (like McDonald's) have managed to exempt themselves from the double-delivery system [for the obvious reason that customers hate getting cold food!].


I did pizza deliveries back in my youth. Got 2-4 dollars plus tips for every trip. The 2-4 dollars essentially covered vehicle costs and the tips were what I made.

Thing is when I did we kind of optimized for the trip costs / payout. Deliveries only happened (for free) for orders over a certain value. During peak times (around dinner) we’d also batch orders. I’d often go out with 3 or 4 deliveries at a time and come back in about an hour with $30 or so just in tips. The 3 hours or so of dinner rush basically paid for the rest of the day.


For me the best was corporate offices. One order was an entire conference room of people ordering on the company dime with an auto 20% tip. Some times I would get close to $100 for a trip in the late 90s for 30 minutes to unload and load a car. Delivering flowers was also $10 a delivery. I would take off school on “flower holidays” and crush it at nursing homes and stuff. My parents were supportive of my hussle.


New Year’s Eve as well. Happy drunk people tip really really well. Also not sure they always knew how much they were tipping, but hey I was a broke 16 year old, I’ll take whatever is handed to me. I remember a couple of New Years eves I walked away from the night with a few hundred bucks. That was a lot of money for me at that age.


> Driver loses money on the vehicle depreciation much of the time.

This shouldn’t be as much of an issue, compared to the ride share market, as you can make deliveries in a $2k junker for years before it gives out.


You can't expect run a 2k junker on significant monthly miles without incurring costs. You might be lucky, but it's a bad bet. Not to mention the risk that some mechanical failures may cost you your job.

Old cars are definitely cheaper to run (net) than new ones, but they aren't free.


True, although that would probably require a shift to more full time drivers. My last Chinese food order arrived in a new Caddilac.


This seems to be an area where europe is v different from US.

How would it work in the US? Would it work if the app did not take a cut at all? Is there just little demand for delivery in the US?


Agree. I don't think restaurant delivery is crying out for central planning. It's overhead that would be better spent in any number of ways.



Where are these $1.99 burgers? Every app I check has like a £10 minimum order, £2.50 delivery fee, £0.50 service fee, please tip our driver prompt, and jacked up prices compared to in-store.


The delivery fee is $1.99, not the burger.


Sorry, I get you. Yes I agree, that's not a sustainable price to be delivering food all over the place for. The depreciation on the driver's car is probably more than that.




Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: