Network effects are the moat here, much like for ride hailing apps. More drivers signed up on the app means that the service times for customers are improved, and more customers using the app means less waiting time and dead mileage for drivers.
There isn't much of a network effect at all once you hit a certain, relatively small, size which is why national ride-hailing businesses keep pushing the big ones out of their respective markets, and why there already are quite a lot of them competing profits away.
There's an illusion of network effects due to the investors willingness to burn their own money. Otherwise, like the actual transportation industry it's just a low margin business with tons of competition.
Meaningful network effects don't just make your company marginally better as you get bigger, that's true for virtually any business. It has to be shown that the size actually makes you run away with it instead of just leveling off.
The same is true in the fintech industry with all these much hyped neo-banks running on debt. Internationally they all tend to lose to local competitors because knowing your market and getting the details right is more important than scale.
A few things undermine the networking effect: drivers working for multiple delivery/ridesharing companies, and customers using multiple delivery companies.
If Postmates went of out business tomorrow the drivers would just shift to Doordash and Grubhub and Instacart and the others. There is zero moat.
I wouldn't call it network effects. All competitors are predominantly sharing the same network of customers and drivers. Sure, some are loyal to one service but I'm guessing a majority of people will try to get the best deal. If the network participants essentially select for lowest margin service with low switching costs, there is no moat. All services, in theory, should gain efficiency from network scale at same rate.
