Hacker News Reader

> It's arguably the most valuable battleground in logistics - developing a "last mile" delivery network at scale.

But it's also hard to get the prices up enough without people just getting off their lazy asses and buying food.

source: price goes up I get up off my lazy ass.

I don't think they can increase service fee much more. The endgame will be to have enough users and drivers that they don't need to increase prices and can make their share from volume, ads and deals with restaurants. Economy of scale + Winner takes all.

> Their core business - ridesharing - has improved their gross margin and unit economics quarter-over-quarter to the point that operating and discounting losses here are almost non-material (less than $22m total worldwide this quarter, representing less than 1/2 a percent of total ride sharing revenue).

That's pretty good. Seems like there's a huge incentive to just start a bunch of big projects when you raise such giant rounds of money.

So are the unit economics any different for DoorDash, which has raised a ton (from SoftBank, no less) at a $12B valuation?

> from SoftBank, no less

Is this supposed to be evidence of a sound company or a strike against them?

After Wework I really can't tell.

The one area that they're supposed to be good at, mobile telephone networks, they've totally flopped with Sprint.

To be fair, Sprint was doing REALLY poorly, and I'm not sure what could have saved them except a merger.

I remember their quarterly calls spending much more time on the accounting maneuvers that generated the cash to cover losses than other parts of the business.

There's no meaningful moat in the last mile, no economies of scale that you gain by being in more cities. You will always be subject to a race to the bottom in every location, as soon as you stop pouring in capital the competition you drove out will drive right back in.

It's a mug's game.

If you gain dominance in a city, you've got a real advantage. The reason is because you can aggregate demand and supply. Once you have enough scale the problem turns from "Get one person to deliver one meal to one location" to "Get this person to deliver these 5 meals from this restaurant to these 5 apartments". Suddenly, you're doing 5x the deliveries for only 2x the cost of transportation.

The game becomes a matter of how efficiently you can map multiple orders to a single delivery driver and the more demand you have the more likely you are to be able to allocate multiple delivers to a specific driver. The more orders you can allocate to a single driver the more margin you have.

I'm not completely convinced. The argument is due to economies of batching delivery. But restaurants can batch their own deliveries too. That line is the "inventory / order interface" and there's no universal optimal point for it, it will vary according to circumstances. Anyone who sells expensive food may choose to optimise for service, which means delivering themselves. Those who optimise for volume will eventually cut out the expensive middleman.

Adding to which, the most profitable markets are the dense cities where there are already strong incumbents, which goes to pouring in capital as a substitute for building a business organically.