Hacker News new | ask | show | jobs
by alaskamiller 3316 days ago
You look at it as a restaurant problem but it's really an ecommerce supply chain problem.

By implying restaurant you're still thinking about back of the house and front of the house. That paradigm doesn't exist for the likes of SpoonRocket or Munchery or Sprig.

They operate out of commercial commissary kitchens that produces products that needs to go into a distribution network to people funneled in from marketing. The bet is that function isn't linear, rather it can be scaled up, especially when prodded along by all the state of art and best practices in ecommerce marketing techniques.

These guys were trying to play the Amazon game, but instead of a 2 day shipping window, it's right now or in a few hours. And instead of a shipping hub to your door steps, it's down the street to your doorsteps. And instead of elastic goods that can wait for awhile before you commit to buy it's a need to satisfy.

A successful restaurant feeds your craving. A successful supply chain fulfills your need.

The real challenge is just the fact at the end of the day there's a fixed cost to fulfillment no one can figure out how to shake.

Despite all the twisting in the labor relationship (avoiding W2s or stealing wages), it still costs $15/hr to deliver up to four $10 items that costs $3 to make. Until that goes down (with automated tincans on wheels) then merely sale of food items won't support these bets.

Except weed and alcohol. Those are the only things that have the margins and demand to support an immediate delivery network.

4 comments

Being a delivery-only restaurant isn't trying to "play the Amazon game" - it's trying to be a restaurant that only delivers.

OP is right here that the substitute good for Sprig was ordering delivery from a restaurant. Typically when you think about ordering delivery, you instantly start thinking about your favorites, meaning this industry is hard to scale because it requires building trust.

Weed and alcohol may have monopolistic characteristics brought on by strict licensing requirements, but people typically don't consume these on a daily basis so the lower delivery quantities negate the higher margins.

in the UK a few restaurants are experimenting with this and years a go one of my local Indian restaurants also had a take away only kitchen.
What even is a delivery-only restaurant? A restaurant is literally defined as a place you go to eat and pay. A fine dining experience is composed of front of house and a back of house.

So Sprig is not a place you go to eat. There is no front of house experience. It's a thing that makes something to deliver to you. That's a fulfillment service.

Here's something to think about, most restaurants don't offer take out or delivery services. Because it undermines their value. They want you to sit down and enjoy in their decor, talking to their staff, eating their food, and most importantly they want exacting control and consistency. That indeed is how you build up trust and repeat visit.

But you're also conflating two behaviors here.

When you think about ordering delivery it's because you're trying to satiate a craving. You want that one thing form that one place. Delivery is an augmented service to the restaurant. It's how DoorDash can exist. But it's also why In 'N Out Burgers sued to not have their food be delivered. And it's also why the bottom half of the DoorDash's platform are not good experiences, it's a restaurant being pulled to be something it's not.

But that's not what Sprig is. They don't offer the same thing every day, they are relying on you to trust the top level branding alone and by browsing the options. It's a production facility with a delivery network.

When you want to your favorite meal delivered and when you want to order from Sprig or SpoonRocket or Munchery are coming from two different places. The former is hacking a restaurant, the latter is purchasing products.

Then lastly, it's not about licensing requirements for weed and alcohol. It's that they're inherently different products whereby immediate fulfillment is possible aside from transporting a human being.

> What even is a delivery-only restaurant?

90% of pizza joints. Yes, they may have seating area, but it's greasy, uncomfortable, and small. The overwhelming majority of their business is take-out, and delivery.

Every single hole-in-the-wall place. They either deliver, or they get you to pick up your own food.

Food trucks.

Having no front isn't some amazing innovation that will afford you margins that VCs and computer programmers expect.

Yes, optimizing your logistics and supply chain in the restaurant business is a huge problem. It's also a problem that successful restaurants, by necessity, are already really good at solving. You're not going to squeeze much blood out of that rock.

This is the reason it will be very difficult for anything like the Sprig concept. Food delivery is insanely competitive and it's a race to the bottom to see who can stay in business on the smallest margin. That's usually the place that operates out of a low-rent storefront and employs family members at below-minimum wage, not a glitzy well-funded commissary with acres of stainless steel and new equipment.
Restaurants and other food-related businesses tend to have an extraordinarily high rate of failure. They're really hard to sustain long-term. So the fact that a food business failed is not remarkable; that's the usual, expected outcome.
And 100k/yr developers. I never understood the excitement around food delivery start-ups. There's nothing like broiling a nice steak with onions. It's like Americans have forgotten how to cook.
>what is a delivery only restaurant?

Some restaurants don't have any seating area. Think Dominos Pizza.

>When you think about ordering delivery it's because you're trying to satiate a craving.

Or you just don't feel like cooking. Also you might want anything random - something different you don't dislike. And you don't feel like walking to a restaurant and sitting there alone waiting for food + paying tips etc.

Sprig/SpoonRocket/Munchery/doordash/Grubhub/Eat24/localpizzashop.com/any local place that is walkable distance and offers takeout are effectively direct competitors for me.

The definition of a restaurant is "a building where people go to eat", if you can't eat there, it's not a restaurant…
Are we really going to start splitting hairs about "delivery place" vs "takeout" vs "restaurant"? It's clear what the OP is talking about; cooking food and delivering it is not a novel concept.
On the other hand, the article says they started "experimenting" with serving food to walk-in customers at their kitchen. Which would mean at some point near the end they became a restaurant by the definition you are using.
> with automated tincans on wheels

I think this was the most insightful part of your comment. When cars are self-driving, there's no reason they'll look like cars any more. There will be people-carriers, but they won't be the only vehicles on the road when they don't have to carry people any more.

Instead, what about small cargo carriers, like bike messengers without the bike? Imagine basically coolers on wheels, zipping along with cars. Why have a full-blown car or motorcycle when you can have a tin can on wheels deliver stuff?

> Why have a full-blown car or motorcycle when you can have a tin can on wheels deliver stuff?

For one reason, to prevent people from stealing the goods. I used to deliver pizza. It was pretty much a given in certain areas that if I was not very careful about locking my car that I'd get pizzas stolen out of it while I was parked making a delivery.

A self-driving igloo cooler is just going to get kicked over and emptied.

Those are here in silicon valley already.
Really? What do they look like?
An igloo cooler on wheels, more or less. They are sized for sidewalk travel, not street.

[edit] Here's a brief video that shows what they look like.

http://www.cnbc.com/2017/04/21/robots-are-now-deliving-food-...

Huh, that looks very interesting, thanks. For some reason, I keep thinking those robots look like they're one wrong turn away from being dismantled by thieves in seconds...
I'm sure you're right. This other video claims it isn't a problem, but (simply my opinion here) it watches more like SV bubblethink than something based in reality. We'll eventually see, one way or the other.

https://www.youtube.com/watch?v=UPZwnc_Lk2M

it still costs $15/hr to deliver up to four $10 items that costs $3 to make

It seems like Sprig should have had a major advantage over delivery restaurants though - they didn't have to travel back and forth to the restaurant. With a fixed menu of a few non-heated items, they (presumably) loaded the truck with enough to satisfy the demands of the day and then their only problem is Traveling Salesman.

I don't understand why this isn't a killer model. Delivery restaurants make a living, right?

I'm inclined to believe that the parent comment is right, that the expectations were just too great. Even if the food delivery business was much more profitable than the average restaurant, they still had to support a very expensive VC-built technology team.

Also, I just wasn't that excited about the food. I ordered from Sprig a few times; it was just sandwiches and salads. You can get those within about 50 yards from any office downtown. I'm guessing that delivery restaurants make a living because their food is compelling enough that people are willing to wait an hour. Maybe remarkable food doesn't keep all day on a truck?

I find it hard to believe that having to occasionally return to the restaurant to pick up more orders represents such a large percentage of a restaurant's cost.
The cost of the delivery driver is amortized across the total number of deliveries the driver makes. If you can double the number of deliveries, you've halved the cost of delivery.

Having to return to a fixed point every 3rd or 4th delivery seems like it would add up.

You're assuming the delivery driver can be kept busy throughout the entire shift. I doubt there's many places where a driver could do a single pickup and then be kept busy for 6-8 hours straight.
Can you actually double it? I am not so sure. Even if you could, what percentage of cost does that actually represent?
We do see a lot of success with alcohol and cannabis delivery services here at Onfleet. "The vices" have high enough margin to support this model. Food delivery can also work, but razor thin margins so to make the food and operate the network is tricky.
Weed and alcohol don't expire in the same way, and all the work making it is already done.