If the service cannot compete on it's own merits, and is being propped up by an unrelated business, how competetive is it actually?
If the service cannot compete on it's own merits, and is being propped up by an unrelated business, how competetive is it actually?
This is only true if external actors redefine "the service" to be something other than what millions of customers voluntarily purchase (Prime). For example, I subscribe to Prime because I get free/better shipping on goods as well as video. I don't know what % of my subscription fee I would allot to either, but fortunately I don't have to do that.
Disney makes a lot of money in theme parks etc.; Walmart makes something like $42B annually in healthcare; Costco sells cars & travel packages; Kroger sells gas (petrol); Apple (also a competitor) makes most of its money selling phones.
Why leave all those businesses intact and target Amazon?
That just doesn't seem internally consistent.
The Disney corporation is built around exploiting the rights to these media IPs to the fullest.
In contrast, Amazon is the online version of Walmart that can also sell you computing time and makes feature films which are just a perk of their "fast shipping" club.
There's a pretty clear distinction here, even with one of the other mega-conglamerates (disney) that probably owns too many different content production units.
Where do we draw the line? Should Apple be asked to spin off iCloud because email, photos, documents should all be independent services?
Note that neither Prime Video nor iCloud are monopolies, which would bring different considerations into the matter.
Businesses are allowed to do more than one thing. I'm not sure when it became popular to try to strip down businesses to a single SKU each, and it's not obvious to me that consumers (or anyone) benefits from pursuit of that goal.
Amazon's whole retail business started by selling paper books. Extending that to kindle and audio books doesn't seem so outrageously "unrelated". Its more of an ecosystem which has legitimate benefits to the consumer. And once you have audio books, music and video doesn't seem like a huge stretch.... etc etc.
Apple is the obvious example of this. Do you think every single service apple offers should be broken into a separate company? Apple pay, apple music, icloud all seem to be "unrelated" but the consumer legitimately benefits from these things being bundled into a cohesive ecosystem.
Breaking these things into individual companies which has to be profitable based on just their one product would lead to higher prices and worse user experience, IMO.
For non-consumer facing products, the case is clearer. AWS / GCP could be split out nicely. The ad networks of meta and google could be independent without hurting the consumer at all.
Prime certainly feels more like a bundle of unrelated things, with the streaming service, and cloud music/photos/ebooks/etc. stuff no one uses, and whatever Twitch benefits they have now. I guess Costco might actually have some stuff like that, but I’m not aware of anything other than warehouse access, some online shipping benefits, and the travel/car/“Next” shopping.
People may not read books as much as they used to and many still read physical books but lots of people use ebooks.
But I could be wrong about it being rarely used.
You wouldn't get magical efficiencies by combining an oil company with an airline, for example, even though oil is a key cost component of air travel.
The same goes for compute resources and ad sales - both are roughly commodity services with lots of competitors. Youtube gets some efficiencies from vertical integration with both, due to incentive alignment, but also gets a lot of inefficiencies from the added complexity and management and the need to keep up with competitors in the commodity part of the business, which can be expensive.
The semiconductor industry is a place where you would think you would see an obvious efficiency to vertical integration - designing your process technology specifically for your chips. In practice, vertical integration between the chip design and manufacturing sides is almost gone, with a few notable exceptions. This is because it is actually not more efficient to be integrated.
Technical infrastructure seems to be similar to me.
I would consider this unlikely, for reasonable definitions of "significant." Costs to license IP and develop content are likely the dominant term in expenses.
> what kind of deal would they be able to get from the newly spun off AWS?
The same as everyone else, or it was a re-org and not a breakup.