65 comments

[ 3.7 ms ] story [ 156 ms ] thread
This is another instance where investors step in and ruin small businesses in an attempt to achieve a monopoly position by building a platform for an existing type of service.

As illustrated by the following snippets from the article:

> Restaurants are struggling to adapt as they are aggregated by DSPs

and

> No DSP has been able to prove consistent profitability with their existing business model.

An alternative take is that investors are subsidizing the industry, benefitting both consumers and restaurants, at least in the short run.
Nah, this is investors destroying an industry through subsidies. If subsidies make all sustainable businesses get outcompeted, then what happens when subsidies end?
Other people step in and start the sustainable businesses again.
You may be right, but you're making a very strong claim. It's not obvious to me that it's correct.
I think those snippets demonstrate that asking teenagers to take things over to the neighbor's place does not require a global logistics network. If you insert one, you're just handing another slice of a thin pie out while losing control over that part of your business.

And yeah, it is rather annoying to see too much money chasing bad ideas screwing up local businesses. Uber is a cautionary lesson; apps are not magic pixie dust and the cut has to come from somewhere.

(comment deleted)
>> investors step in and ruin small businesses

Food delivery platforms will get centralized. If not by investors, By Amazon. By someone.

Why ? Because consumers want that. It's more convenient. And probably network effects apply here.

With food delivery centralized, Dark, big centralized kitchens will win over restaurants(for most popular meals). They're just cheaper. It's just specialization and higher volume.

Long term, making delivered food isn't a good place for a small business. And that'll happen with subsidies or without.

BTW: in the future, in most industries, a small business will only survive by making rare and unique things.

I'm not quite as pessimistic on the food delivery market as you are and would be interested to hear your/others' reactions.

>Food delivery platforms will get centralized. [...] Because consumers want that. It's more convenient. And probably network effects apply here.

I agree consumers want all restaurants in the same place. I'm less convinced that that implies platform consolidation because every party has incentives towards multi-tenanting to ensure access. Both restaurants and end customers have incentives towards multi-tenanting to reach each other and for a market-making DSP to demand exclusivity, they have to undermine the core thing both sides of the market want (access to supply/demand). A16Z has a good discussion of multi-tenanting as a limit to the economic value of network effects here [0].

> With food delivery centralized, Dark, big centralized kitchens will win over restaurants(for most popular meals). They're just cheaper.

I agree there are cost efficiencies to dark kitchens and the delivery market is likely to trend that direction (cheaper real estate, leaner front of house, less decoration work, etc). It's not clear to me, though, that there are worthwhile economies of SCALE to be had here (at least in the core urban market), or that they're necessarily winning.

* Ingredients are low-margin, with prices dominated by transportation, so bargaining power over suppliers is limited.

* Delivery time matters. That's a small pro for large kitchens because you can more accurately forecast demand/staffing, but also a medium-sized con because 5 locations spread out in a city with 20 units of capacity each are more likely to be close to demand than 1 location with 100 units of capacity.

* Cooks aren't fungible, which makes it hard to capitalize on scale without consolidating similar restaurants. The guy who makes my favorite lamb vindaloo is a TERRIBLE sushi chef (in fact, he doesn't even make a very good a saag paneer, imho). If he were trained, he could probably learn to make a mediocre pad thai, but I'd probably still be happier ordering from a known thai restaurant. With 70% annual turnover in the restaurant industry[1], you're going to need to pay for a lot of training to capitalize on the more predictable staffing by supporting a large (i.e. multirestaurant) menu without sacrificing quality. If you plan to support a small menu at scale (e.g. combining a NY borough's worth of Indian restaurants into one giant Indian kitchen), the demand you'll need to support it will be spread out over a borough, so you're likely to have difficulty maintaining service levels.

* People have different tastes in food and there's value in variety. I probably have a neighbor who really likes my restaurant's saag. If you merged the two Indian restaurants near me, had an expert pick the best recipe for each dish on their menus, then taught those to the combined kitchen staff, either my neighbor or I would be disappointed in the saag.

>Long term, making delivered food isn't a good place for a small business. And that'll happen with subsidies or without.

Without subsidies, this ultimately depends on whether it's harder to make good food or to build awareness of said food. A restaurant can hire a delivery driver to transport food, which puts a cap on the cut a DSP can demand from them for the delivery portion of their service. It's slightly less efficient, but trivial to implement. The expensive/hard part is getting people to place orders before your lease runs out. If there are multiple DSPs willing to list your food in order to sell their own delivery services, that seems like a recipe for lower start-up costs, which would tend to favor pop-up restaurants/Kitchen Incubators/iteration at small scale relative to the classic sit-down storefront restaurant model.

[0]

Great analysis. You've given me some food for thought.

How did you get such deep understanding of that industry?

> How did you get such deep understanding of that industry?

I'm not certain I have.

This is just how I'd expect it to work based on the arguments in the slides, my own dinner decision habits, a summer working at Chili's, and a consulting/pricing professional background.

>People have different tastes in food and there's value in variety

Which is something that scale helps with. If one person wants Thai, the other Italian, the other Greek, and the last Indian a mega kitchen could fill that order.

>* Ingredients are low-margin, with prices dominated by transportation, so bargaining power over suppliers is limited.

With 100 orders an hour going out from a kitchen you're more likely to be able to combine deliveries. That's an easy way to double your profits since transport is such a big percentage of your overall cost.

>Cooks aren't fungible, which makes it hard to capitalize on scale without consolidating similar restaurants. The guy who makes my favorite lamb vindaloo is a TERRIBLE sushi chef

Eh, they're pretty fungible. Maybe not the chef designing the recipes but certainly the guys chopping vegetables are.

Yes but that is not the point. The point is that it is unfair that just by collecting a huge pile of money from investors you can grab the profits in an entire industry, considering also that creating a food delivery portal isn't exactly rocket science. I.e. literally anyone could do it.

I think governments should deeply think about what is happening with platform economies, and perhaps they should strongly regulate them. An extreme view could be that running platforms is a task for governments.

Also, DSPs are now using restaurants to eventually bootstrap their dark kitchen concept, which is unfair. It's like Uber using human drivers to bootstrap their driverless cars business.

Business is not about fairness. Largely, it never was.

It was always about power: the power of local monopolies, the power over workers, the power of unions, China, laws like the minimum wage, capital, etc.

(comment deleted)
When do the DSPs start to vertically integrate - setting up their own dark mega kitchens, with generic styles like 'Pizza', 'Thai', 'Chinese', 'Indian'? These could appear as different restaurants, but in reality operate out of the same facility.

In this way the batching could be done even more efficiently, and orders would be coming from one mega-kitchen, with huge economies of scale.

Uber Eats is sort of experimenting in this area with "popup" restaurants, only the cooking is done by their existing suppliers. For example, there is a local Puerto Rican restaurant that has a "pop-up" that offers a few Asian dishes. You can't walk into that restaurant and order the Asian food.
There's a service in Indianapolis that does exactly that! https://www.clustertruck.com/

They actually license the recipes from local well known restaurants, and then cook them in their own kitchen and deliver them.

They already do - I wish it was more transparent in the app as it's currently ruining the experience.
“Delivery Service Partner (DSP)”

I don't know why but i kept forgetting what that meant and seeing the acronym over and over bothered me. Future looks bright in this area, looking forward to seeing what happens

Best shared content on HN in a long time. I love the thoughtfulness and thorough analysis of the winners and losers of each strategy.
I used to live in Dublin until recently and coming from a country (Greece) where food/coffee delivery is free (apart from any tip), fast and ubiquitous, I was taken aback by the fact I had to pay a substantial delivery fee.

The upside was that driver safety seemed to be taken more seriously (it's a cutthroat business in Greece and drivers are pushed to their limit, accidents being rather frequent in their line of work).

How do the Greeks afford to include delivery cost in food prices, a business usually with minuscule margins?

How much does a cup of coffee cost if the price includes delivery?

And does every food item include enough cost to support delivery independently? How do they amortise the delivery costs across an order?

The margins were small largely because a huge chunk of it goes into prime estate
It's seen as part of the standard business expenses like utilities and rent. Of course it squeezes their profit margin but what can they do? Competition is fierce. What most joints do though is to require a minimum price per order to allow delivery.
Question: How hard would it be to write create open source components to run a DSP collaboratively, e.g. with the power asymmetry created by aggregation

> aggregated by DSPs

removed?

The same argument was made for an open source Uber, and I think it's a noble idea, but it appears the problems of routing and competition are complex enough that it just doesn't work.
I'm not aware; are there any examples of super apps in the US? If SnapChat added Dominos delivery, would that be a satisfactory hypothetical?
Reading between the lines, the only real opportunity for a sustainable moat is the dark kitchens and order batching. This has been tried and failed a few times already, my guess is the secret ingredients are scale and marketing.

Delivery from traditional restaurants is just too low margin, historically it was done by only some types of restaurants at break-even or even a slight loss to increase demand, now that people are used to paying $6-10 premium, once the music stops and the VC money stops flowing, the restaurants that stay in business will easily adapt.

The tools required for a marketplace that connects people to existing restaurants and a "marketplace" that connects people to multiple "restaurants" that are in reality all from the same dark kitchen are pretty different. The former only requires staff, the latter requires real estate, equipment and product in addition to staff. It's essentially building a restaurant with a tech component, not building a tech product.
I skimmed through the slides and didn't see a single word allocated to the problem of the waste, pollution, and health effects generated by these services or any costs involved in managing them. Perhaps the figure of $365 billion is a convenient fiction, as is so often the case when modern young "entrepreneurs" see an opportunity to upend the underpinnings of society for profit.

Food isn't a problem we solve with megacorporations and apps and technology. Food isn't supposed to be optimized and analyzed and held to KPIs. Food is best when communities of people come together to participate and interact and nourish one another. All these services do is enslave and silo us away from each other so that the individual's essential needs for nutrition and community can become a profit stream in a corporation's ledger. They would build yet another layer in America's enormous neurosis around food. No thanks, Google.

That.

I'd like to see a future of food delivery which is less food delivery, more time to spend at home, more cooking and healthy choices.

But of course, as an endeavor whose main end goal is money, the future of food delivery has to be "more, more, more".

Let's not throw the baby out with the bathwater.

Food delivery is a good idea. Myself, I'd like to see less time spent on cooking. Except few special cases, cooking is a chore for me, not pleasure.

What isn't a good idea is inherently unsustainable, VC-subsidized companies lobotomizing the industry in order to strip-mine it for value long enough to get flipped, leaving a shell of a market sector behind.

>Let's not throw the baby out with the bathwater.

If anything, I'd like to throw out more of the baby: e.g. people learning to grow their own fruits and vegetables, and make more of their own food, not just cooking.

Sure. I don't see why the two can't coexist.

It's also a matter of life stages. I know I want to have a vegetable garden of my own, either vertical in the city, or regular one in the country, depending on where I end up in the future. But today, I very much enjoy the ability to order a dinner or a pizza whenever my wife and I don't feel like cooking after a full day of work.

(comment deleted)
A bunch of delivery-only restaurants have opened up in my city. Most of them are helmed by young chefs who trained or lived abroad and are bringing interesting new cuisines and dishes to local people. It's great for the local culinary scene and the food is often amazing.

But my word, they've somehow come to think that since they are delivery-only, they have to give customers a "dining experience" by overpacking everything. Every order comes in a fancy cardboard box, every order has multiple sets of disposable cutlery. Every little dish is packed in its own plastic box. With some cuisines (like Korean), you end up with over a dozen boxes for a single meal.

The amount of packaging waste delivery creates is a little sickening. I've taken to ordering from restaurants that throw the entire order into a cheap cardboard box and skip the "branding" altogether.

Okay, but how is optimization of delivery not going to help address waste, pollution, and health effects. We have so much abundance in Western nations that we fail at times to understand how much optimization through technology, apps, and whatnot, will still provide benefit.

From reducing wasted food by ensuring only what is needed is on hand and in the channel. Covering pollution by reducing trips made by many to fewer delivery groups. To health by offering better choices as part of the service, new menu options never considered because they were not previously available.

The food abundance did not come way of the government but instead by corporations and individuals trying to improve their position in society all the while providing services and goods with improved the lives of others. If anything tariffs and food regulatory groups make food more expensive by preventing our access to other producers and products.

(comment deleted)
If you live on takeaway food, yes. I work for a DSP company partnering with Just Eat and I also do a lot of home cooking with local non-intensive produce.

Food delivery is way too expensive for most people to cater for every meal. A Friday night treat after a rough week has its rightful place in the world.

As for pollution, food is increasingly being delivered by bicycle in many cities here in Europe.

Food isn't a problem we solve with megacorporations and apps. Food is a problem we solve with agricultural production at scale and shipping and logistics. Food is best when it's nourishing and nutritious and where people are in sufficient quantities to cover everyone's needs with foods they are familiar with.

This is just my own wildly unreasonable opinion, but perhaps we could consider the possibilities to be realized by shelving the romance of participation and community and unity until we've got the logistics solved. Rather than trying to tell people how they should relate to food, maybe we should focus on making sure everyone has enough and that an obsession with hyperlocalism isn't setting the world up for more famine.

Perhaps then people can choose how to relate to food and how to nourish their communities. I'm certain that many people will choose exactly what you describe.

The logistics is solved, or solved enough at a price point the HN class of people can afford, hence the focus on the experience and community of it. The upper classes want fancy "experience beer" you go to some brewery bar to drink with your friends instead of just "decent, beer" that you drink from a can while sitting in a lawn chair on your own property.
Isn't food the underpinning of most cultural traditions? It seems to me like society (as defined by generations of civilization) has already determined that food has value for it's experience and community aspects. Maybe it's more than just an "HN class of people" thing and you could examine your own perspective?

[1] The cultural significance of food and eating - https://www.cambridge.org/core/services/aop-cambridge-core/c...

[2] How Food Shaped Humanity - https://www.smithsonianmag.com/arts-culture/how-food-shaped-...

[3] Understanding Culture: Food as a Means of Communication - https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&c...

Food, and its production and consumption, is definitely the underpinning of many cultural traditions. Today, many people have the opportunity to choose how bound they are to such traditions.

Some people prefer to do their own cooking, at home, for themselves. Others prefer to have every meal prepared by a professional in a restaurant. Still others prefer to have most meals delivered. Some like to mix and match as is convenient.

Unbound from tradition, people can make their own choices, and decide for themselves how much they value traditional romantic attachments to home cooking, participation, and interaction in their food experiences.

I think you are missing my point. At it's core, food is simply fuel for the human body. In practice, it is a common thread that unites different cultures and generations. I am suggesting that we should examine why food plays such an outsized role in our cultures. Is it a part of our social nature, oral traditions, basic necessities, etc.? Is it something that can be "disrupted" like e-commerce or some other man-made industry?
I think you have made a cogent point. Food is common and an essential part of how every human culture defines itself. I think you have raised a wise question about how disruptable food really is.

It might be worth considering that human history - recent and otherwise - might provide an answer. Human food culture is definitely a thing that can be disrupted like any other man-made industry. Restaurants, delivery, central kitchens, self-service grocery stores, relgious food rituals, and more were all at some point major disruptions. It's perhaps possible that the critical and unique centrality of food to culture you correctly point to might not inure it to disruption.

You're absolutely right that there is a discussion to be had here. Might it be better understood as a discussion explicitly about social rituals, oral traditions, and uniting humans across generations?

==Might it be better understood as a discussion explicitly about social rituals, oral traditions, and uniting humans across generations?==

This is a good point and you are likely correct. I think I was internalizing some things I've read recently on HN that haven't sat well.

How exactly are waste and pollution made worse by people delivering food to you versus you driving to the restaurant and buying it?
Because people are more likely to just cook at home, if delivery weren't an option?
Do you have data to support this hypothesis? Bear in mind that various forms of takeout food and restaurants have existed for thousands of years in many cultures.
taxis have existed for over a century, but ride-sharing apps make "taxis" more popular due to increased availability/affordability/convenience.

this is all speculative, i've no data :)

(comment deleted)
* [Kitchen incubator - Wikipedia](https://en.wikipedia.org/wiki/Kitchen_incubator)

* [Communal apartment - Wikipedia](https://en.wikipedia.org/wiki/Communal_apartment)

* [Communal dining - Wikipedia](https://en.wikipedia.org/wiki/Communal_dining)

* [Soup kitchen - Wikipedia](https://en.wikipedia.org/wiki/Soup_kitchen)

* [Factory-kitchen - Wikipedia](https://en.wikipedia.org/wiki/Factory-kitchen)

* [Perpetual stew - Wikipedia](https://en.wikipedia.org/wiki/Perpetual_stew)

* [Bhandara (community kitchen) - Wikipedia](https://en.wikipedia.org/wiki/Bhandara_(community_kitchen))

* [Coliving - Wikipedia](https://en.wikipedia.org/wiki/Coliving)

* [Campus Kitchen - Wikipedia](https://en.wikipedia.org/wiki/Campus_Kitchen)

* [DC Central Kitchen - Wikipedia](https://en.wikipedia.org/wiki/DC_Central_Kitchen)

* [Karma Kitchen: Growing in Generosity](http://www.karmakitchen.org/)

* ...

* ...

* ...

* ...

Yeah, markdown or something like it would be a nice flourish. But I guess HN kinda likes to keep it simple instead. shrug (hey -- that's some formatting right there)
"Ethics aside, they can..."[1]

I expected nothing less from somebody who worked for Uber.

I mean sure, investors will (pardon the pun) eat the ideas up. There is, after all, money to be made. But the overall state of our industry, where people feel very comfortable discussing business models with a prefix of "Ethics aside" is depressing as hell.

Let's not even mention the problem of externalities, of impact on rural communities, of the immense problems that consolidation will bring, or of the exploitive nature of the "gig economy"

[1] Slide 25: https://docs.google.com/presentation/d/1v9ifaW1Sxy_oVUtzn63B...

sort of humorous that they included a cartoon which is mocking a part of their proposal
Forgive me but as someone who has a fridge, microwave, stove and oven, delivery does not appeal to me.

There is a whole frozen pizza in the store that sells for 3 dollars, and a slice of pizza from the 7-11 that goes for 2 dollars. Pizza delivery is the one food where delivery has matured, and even today it's a very limited buy.

When do people do it?

Parties -> Almost any demographic

Workload is too high to go out -> Students and younger adults

Going outside is extremely uncomfortable and/or dangerous -> Primarily older

Unmotivated and lazy -> A small fraction of any demographic

The thing that can finally drop prices as we know it is the quadricopter. A self navigating delivery channel has at least reasonable chance to bring the price well into tolerable levels for most americans.

You assume the unmotivated and lazy are a small fraction of any demographic. I have an unqualified feeling that it is a big portion, and that it is growing. I hear too many references to netflix tv shows, and a lack of conversations that spark my curiosity. I hear of people not sleeping to catch the latest episodes that aired in another country first.
Not everyone enjoys cooking. There are also some very motivated people that simply would rather spend their free time in other ways.
Every time I see a delivery bike out with a huge backpack full of food, I feel as a society we've gone a bit wrong.

1) The delivery rider/driver is making nothing 2) We are packaging every piece of food and sauce in single use plastic 3) The restaurants are losing money (though this is being addressed somewhat) 4) People aren't leaving their homes for the 5 minutes it takes to walk down the street and support a local business.

I really hope this is a trend that goes away, but I doubt it will.