I mean I think it can be, given that there's a $20-$30 markup per order, but it definitely requires some particular circumstances.
The bigger thing is that if you're thinking of Lean manufacturing principles or Theory of Constraints, where the goal is to get the noise out of the system, you would never maximize a business the way that Uber Eats and Doordash are. It'd instead be something like the following pipe dream:
- We sell "pizza delivery as a service" to businesses who want it -- at first we're trying to partner with Target and Walmart, etc. Our value proposition to them is "hey Amazon is eating you alive, how about we help you offer ultra-rapid delivery and the people you're working with basically become fellow trusted Target employees?" (Of course, we'd be happy to double-dip -- ideally we'd convince Amazon that we're a cheaper way to reach their customers than their own delivery drivers. But that's a really hard sell.)
- The company rents delivery people from us per day, we provide the delivery network. If you need to "burst" we can provide extra folks to augment at a premium. Ideally someday we'll get some lucrative Amazon contract where the Amazon warehouse, but to start with we're doing laundry delivery and other odd jobs. The company maintains their brand; we're just a courier service. You don't go to the FedEx website to order something, you select FedEx at the end with "how would you like this to come to you."
- On the side, we do sell a service of "we will give you a new pickup/delivery portal web site that works with our delivery service" to smaller businesses so that they can get started with us.
- We are a courier service but we only operate in cities where we can sufficiently average out the volumes needed. Amazon, UPS, FedEx can be the kings of the countryside, that's fine.
- The company, assuming food, sends their delivery-person right when the food is cooked, they are our employee and package it to our standards and bring it straight out to our delivery hub. In general they will not be going to the final address. We use this as a process buffer to maximize our throughput for the people actually driving to the home addresses, we can make sure that certain people get to know neighborhood X better, etc.
- The shipping price should come in two tiers, give people a generous shipping discount for advance orders where we can have more opportunity to batch and optimize.
- Because the company pays us (and the user pays them for "shipping"), under no circumstance does the user tip; that workflow is just too indirect to sustain tipping.
- Which means we have to do a 180 on how we treat the delivery folks; ideally they'd be full-time employees and we'd proactively unionize them.
That last bit sounds like suicide but really there's a ton of noise in terms of all of this "well the question of who are our delivery people is fraught, they can drop off the map at any time" and it's like, no, I want to be able to say at the central hub, "here's the route you're about to follow, review it while you wait 2 more minutes for Tina to arrive with order 6AF12B, that's your third one, we've got the motor running for you and you know these streets better than anybody," and under no circumstance is that person saying "eh, I have a party I want to go check out, I'm just going to go with these 2 orders and leave Tina in the lurch." And, we can sell it to city-goers as "this is the ethical way to deliver, the union makes sure that we maintain the cars and that we pay as well as we can."
And the rest of it is making the profit back up in volume.