Hacker News new | ask | show | jobs
by notyourday 2513 days ago
> 1) Ghost kitchens that only serve via delivery apps. You can save on rent by setting it up in an area with no foot-traffic. You can also get economies of scale.

You can't. This is the kind of a mistake that all first time entrepreneurs make. Delivery costs too much money. Logistics of getting your labor force to your kitchen is a nightmare.

> 2) Restaurants with dedicated entrances for delivery drivers. Baar Baar in NYC has a separate place where delivery workers can wait. In other restaurants, they have to wait in the limited dining area. One hopes that the workers still get to use restrooms & get some water, etc.

This was tried in NYC during first dot com. Remember purple shirts? They had a competitor with green shirts. That's the unproductive spaces. Rent hikes wiped those out.

Edit: Blind downvotes are laughable. Go walk into your local pizza place and talk to the owners. They will tell you why they are slowly dropping off all the apps. They are giving equivalent of a 20-30% discount in fees for the area they already deliver in.

Ghost kitchens are mostly scams. The ones that aren't are ran out of commissary commercial kitchens that cost an arm and a leg. The "ghost kitchens" are mostly not really legal, sometimes ran out of a house, marketing same stuff under different names. Most of the restaurant overhead for small places does not come from the rent of the floor space. It comes from the rent of the kitchen and compliance with the health codes. And, unlike say taxi commissions, health departments take no crap, which is why the ghost restaurants disappear rather quickly.

It is the "no one did it before" idea of people who have never ran restaurants. It always flops. If it did not, Union Sq Hospitality would have been doing it for years.

6 comments

I'm not sure why you're getting downvoted. A beloved restaurant near my house is closing down, and the owner is pretty clear about how ugly the economics of delivery are: https://missionlocal.org/2019/05/custards-last-stand-mission...
Whoa, I had no idea the commissions were that high.
I'm not sure why all the downvotes here, either. In the SF Bay Area, several companies have tried to be "delivery only restaurants" and they've mostly all been failures -- some spectacular VC-assisted flameouts, some quiet fades.
Ghost kitchens may be scams but I know at least of one QSR in Manhattan that does them successfully. In some instances they are quite profitable if you generate the volume.

Your commission is also a bit off, perhaps you speak from full service experience?

While true that base comm is 20-30, those numbers are always negotiated down. Grub/Seamless will take as low as 12 from high vol regional leaders while UberEats made some deals that are supposedly really low (McDonalds). I don’t have exact numbers but from my recollection they mentioned it in their S1.

While Danny Myers is quite smart, USH isn’t a barometer for the entire industry as his focus is more on full service.

> Ghost kitchens may be scams but I know at least of one QSR in Manhattan that does them successfully. In some instances they are quite profitable if you generate the volume.

It is a temporary fluke. If Chinese take out operators that are able to do razor thin margins and nearly completely unpaid labor force cannot make them work then no one can make them work

> While true that base comm is 20-30, those numbers are always negotiated down. Grub/Seamless will take as low as 12 from high vol regional leaders while UberEats made some deals that are supposedly really low (McDonalds). I don’t have exact numbers but from my recollection they mentioned it in their S1.

It is a creative lie by omission. Same kind of lie that Grubhub had when it forgot about the $$ it charged restaurants for answering phones that Grubhub setup.

There are maybe a dozen chains (McDonalds/Burger King/Chipotle/Qdoba/Subway/White Castle, etc) that can negotiate $1 fee. Your pizza place, your taco place, your take out sushi joint don't have this ability. If they did then the VCs would not fund DoorDash or GrubHub.

> While Danny Myers is quite smart, USH isn’t a barometer for the entire industry as his focus is more on full service.

Have you heard of Shake Shack?

> It is a temporary fluke. If Chinese take out operators that are able to do razor thin margins and nearly completely unpaid labor force cannot make them work then no one can make them work

They can, and they do. One way some chains achieve this is by raising prices vs. in-store. By offsetting the cost, the economics makes more sense.

> It is a creative lie by omission. Same kind of lie that Grubhub had when it forgot about the $$ it charged restaurants for answering phones that Grubhub setup.

> There are maybe a dozen chains (McDonalds/Burger King/Chipotle/Qdoba/Subway/White Castle, etc) that can negotiate $1 fee. Your pizza place, your taco place, your take out sushi joint don't have this ability. If they did then the VCs would not fund DoorDash or GrubHub.

I personally know of chains that have less than 50 locations and have negotiated ~12-15% comm. I also know of a NYC deli with one location that runs at ~15-17 comm. Its about volume. You can negotiate anything with high enough vol.

Not sure what you meant by "creative lie by omission" or if you were referring to what I said.

> Have you heard of Shake Shack?

Shake Shack IPO'd in 2015. White Myers does sit on the board (chairman), they are different companies. Also, Shake Shack has only recently started going into delivery (check their latest 10-K) they are fledglings going from their own platform (OLO) to working w/Caviar/Dash/etc.

Again, successful, but not a barometer.

> They can, and they do. One way some chains achieve this is by raising prices vs. in-store. By offsetting the cost, the economics makes more sense.

That's not allowed by the contracts. In either event 50 location chain is HUGE. Olive Garden has less than 1000 stores nationally. If you have 50 stores your chain is in the real estate business and it happened to have food as an additional line item.

> I also know of a NYC deli with one location that runs at ~15-17 comm. Its about volume. You can negotiate anything with high enough vol.

Unless you happen to own that deli I'm going to tell you that you are being lied to. No one does enough volume to do it.

> Not sure what you meant by "creative lie by omission" or if you were referring to what I said

Something else is padded. Delivery services won't make any money off the 10% commission.

> Also, Shake Shack has only recently started going into delivery (check their latest 10-K) they are fledglings going from their own platform (OLO) to working w/Caviar/Dash/etc.

The argument was that USG type companies only do full service restaurants. That's simply not the case.

> Again, successful, but not a barometer.

It is because it has been around longer than most of the startups in the delivery space, not to mention vendors to the startups in the delivery space that do not have actual restaurants.

> Ghost kitchens are mostly scams.

They are not. They are just a different model than a standard restaurant.

What I have observed in my area (UK) is that instead of expanding by setting up a new branch in the costly centre of town, some chains work with Deliveroo to open a ghost kitchen in the industrial area.

That allows them to expand to a new town while saving on capex.

> They are not. They are just a different model than a standard restaurant.

They are not. It is called commissary kitchen. Has been around forever. Those are the kitchens that support food trucks. Or the kitchens that support prepared food not made on premises in supermarkets. They don't work for a-la minute production even if a-la minute is just used to load it into driver's gear.

> What I have observed in my area (UK) is that instead of expanding by setting up a new branch in the costly centre of town, some chains work with Deliveroo to open a ghost kitchen in the industrial area

That simply means real estate has not adjusted to it. It will.

A standard restaurant must be located where people will walk in. It must be pleasant premises for customers.

If the target is to work only with Deliveroo or other delivery app then everything can be stripped apart from the kitchen and that kitchen can then be located where it's cheapest since no customers will ever walk in or even know where it is.

That's not a scam. That's an adaptation, an optimisation for a specific business model.

> That simply means real estate has not adjusted to it. It will.

Premium retail space is going to cost more than industrial space that may be located anywhere in a much wider area.

That's how people who do not understand restaurant industry think of it. It is wrong.

Restaurant wants to minimize dwell time, which is a time that the restaurant does not spend cooking food for a-la minute order. The closer the restaurant is to the customer, the smaller is the dwell time if the customer is always available. That's the reason why the restaurant wants to be where the customers are.

Delivery-only restaurants dwell time is always larger than the walk in restaurant unless the delivery drivers are immediately and always available when the food is cooked and the time to deliver the food to the customer is the same as time time to turn it over across the counter ( for the same type of food ) which means the restaurant will always make less money in the identical situations.

The only thing that the restaurant gets from the delivery only portion is marketing when its menu is available to people who do not know about it. And it competes with every single other restaurant available in the delivery order. All that for the 20-30% discount off every order. That's the math. It is groupon, all over again. And the arguments of those in awe of the business model are exactly the same. It does not work because the best slammed take out places have 20-30% operating margins.

Finally, there's a startup called Crave which is going to kill this entire DoorDash/Delivery/Deliveroo type model - it gives restaurants ability to dynamically price their checks via discounts ( think surge pricing backwards ) for a low, flat per sale price. Think of it as Yelp Cash, but more dynamic and less connected to Yelp.

This actually solves the problem of dwell time. By offering a dynamic discount based on how busy the place is the restaurant increases a flow of orders into the business.

Sounds like the aim of "Monday specials" and "Happy Hour" brought to digital age.
Can you elaborate on the delivery costs and employee travel aspect?

Surely by selecting a location that balances low rent versus proximity to residential areas you could lower operational costs while still ensuring staff could get to work and keep delivery costs low? I get that running a restaurant is a complex and risky business, but i dont think youve really addressed why a delivery only kitchen would fail consistently.

> i dont think youve really addressed why a delivery only kitchen would fail consistently.

It would fail consistently because it has a worse dwell time compared to a comparable restaurant that also does take out while providing no real advantages. There are just no margins for error in a restaurant business. If your competitors dwell time is 10 minutes and yours is 30, you are done.

Here's is why the proximity to the residential area is important: kitchens are staffed by the group of people on the lowest tier of the economic ladder. These people use public transportation to get to places, not cars. The industrial zones where the rent is cheap aren't accessible easily via public transportation which means the kitchen staff has problems getting there, which in turn means that it selects other places.

If an industrial area is easily accessible by a car from the residential areas which is what is necessary for it to be a good source for deliveries then the rent in that area is only slightly lower than the rent in the residential area so there's really no win at putting the kitchen there. To get a win on rent one needs to put it 20-30 minutes away where the rent could be significantly cheaper. Except that it means that it takes at least 30 minutes to get to the border of the residential area with the food, plus additional 10-15 minutes inside the service area, which means the delivery only saving money by putting the place in the cheaper industrial area restaurant has the following minuses:

1. Difficulty to get kitchen staff 2. 45 minutes of driving to get from its kitchen to an average customer if the driver is immediately available at the location [ the drivers is not at the location: the driver is probably at the transit somewhere in or around the residential area so the turn around becomes at best ~1 h ]

It's competitors located in the residential areas:

1. Don't have the logistical problem at hiring staff 2. Can get food delivered in sub 25 minutes of driving

The only positive for the outskirts kitchen is ~20-30% at best save on the rent.

Dismissing things in 2019 by referencing that they were tried in the dotcom bubble ... Two decades ago ... Is not a very powerful argument
> Dismissing things in 2019 by referencing that they were tried in the dotcom bubble ... Two decades ago ... Is not a very powerful argument

Cost structure has not changed in restaurants that do take outs.

Delivery services in the areas where they ( delivery services ) could be popular and are popular solve a non-existent problem for restaurants ( delivery ) by eating into restaurant's profit margins. Not only that, by longer-term delivery services are terrible for restaurants because suddenly every restaurant in 1-2 mile radius even in the most high density city becomes a competitor while the margin is lower.

Isn’t the whole point of history... to learn from it?
Are you suggesting that something changed to make this different? If so, what?
Smart phones (the associated app ecosystem, constant IP connectivity, GPS, and stored payment methods) and better computer mapping/route planning systems being ubiquitous.

I remember around 2000 in order to order ice cream for delivery having to call Kozmo (or maybe it was Webvan) on a voice phone and talk to a person like some kind of savage.

It seems like the smartphone and good map data could transform something that was terribly inefficient 20 years ago into something that's efficient enough to be profitable now.

> call Kozmo

Yes, those were the purple shirts! Thank you for reminding me the name.

The thing is delivery is not an issue. In the markets that have density, local places figured out delivery. My bodega will deliver whatever I order. For free. So will four other bodegas in the immediate area. The guy who owns the wine store will deliver for free. So will the taco joint. Or PR chicken shack It will be done by the kids of people who own those places. Or by their relatives. Or by their helpers. Yes, all of them would also try UberEats and Grubhub and Seamless and others. All of them will drop those services after a few month ( all of them did ) because they are in a real cut throat market and they know what starts costing them money.

So we get to the areas with less density, which is the burbs and near rural. Such as, say, Bucks County in PA within a 20 minute drive of Doylestown. There are probably about 100 places total to get food from, including Wawa, take out places, casual restaurants and "upscale" sit down places. The issue there is that general delivery is not profitable, which is why none of the places deliver outside 5-10 minute drive. Their customers have cars and driving 10-20 minutes is not a big deal for a customer. The max time they are willing to wait for food is cook time + round trip time. Delivery services can't compete there because they can't beat the cooktime + roundtrip time ~ delivery is free option