It ought to be staggeringly obvious that distribution matters when there is a physical good involved, and especially so in time-sensitive situations where customer pressure is strong.
There are a number good points made here, some with the author's own prose, and some via quotes first said by others; it packages them up fairly neat but sneaks in a wishful fantasy where some food delivery company abstracts away from "real" restaurants and serves first-party food instead.
Perhaps this may happen, given the long history of grocery store private brands, white label goods, or even "commoditized marketplaces" where hordes of customers hit up sellers who are highly rated on some 'authentic-seeming' metric local to that walled garden's universe (Amazon Star Rating, Hotwire "Recommended %", Uber rating, Facebook likes)... but this is due to customers' price sensitivity once differentiating factors have been removed, and leaves room for players who continue to offer a differentiated product or experience.
Fundamentally, this is where aggregators excel after all: you can get both your "one taco, any taco will do", and your fancy offering, and you're their customer regardless; meanwhile those who branch out into distribution or production will be able to preferentially plug their first-party product but at the cost of having to be responsible for an entire product lifecycle in the first place. Unless you have a huge warchest or that's your core competency, this is a risky call.
Perhaps the lesson is, if you already have a huge marketshare of customers, you can probably win additional people over with trying to lower prices on your offerings (perhaps by applying price pressure on your sellers, perhaps by making your own things), and the sellers you aggregate will be loath to leave because you still hold the crowds. That's network effects.
There are a number good points made here, some with the author's own prose, and some via quotes first said by others; it packages them up fairly neat but sneaks in a wishful fantasy where some food delivery company abstracts away from "real" restaurants and serves first-party food instead.
Perhaps this may happen, given the long history of grocery store private brands, white label goods, or even "commoditized marketplaces" where hordes of customers hit up sellers who are highly rated on some 'authentic-seeming' metric local to that walled garden's universe (Amazon Star Rating, Hotwire "Recommended %", Uber rating, Facebook likes)... but this is due to customers' price sensitivity once differentiating factors have been removed, and leaves room for players who continue to offer a differentiated product or experience.
Fundamentally, this is where aggregators excel after all: you can get both your "one taco, any taco will do", and your fancy offering, and you're their customer regardless; meanwhile those who branch out into distribution or production will be able to preferentially plug their first-party product but at the cost of having to be responsible for an entire product lifecycle in the first place. Unless you have a huge warchest or that's your core competency, this is a risky call.
Perhaps the lesson is, if you already have a huge marketshare of customers, you can probably win additional people over with trying to lower prices on your offerings (perhaps by applying price pressure on your sellers, perhaps by making your own things), and the sellers you aggregate will be loath to leave because you still hold the crowds. That's network effects.