Hacker News new | past | comments | ask | show | jobs | submit login

What I want addressed is how we can bust these companies without making the experience worse for consumers or making things less efficient.

A railroad that'd require a toll every mile because you split one company into hundreds wouldn't be viable. If I had to visit multiple web pages to get the same result as a google search, that'd also be bad.

Amazon online shopping has a "monopoly" almost purely because the monopoliness makes it much easier and lower friction than trying to find an online store that is 1) somewhat trustworthy 2) consistent UI 3) ships in a reasonable time 4) isn't going to disappear (and importantly, already has my shipping and payment information). I regularly buy stuff online from elsewhere, but I use Amazon more often because it reduces the cognitive load by centralizing/standardizing everything.




The railroad example would be solved by forming a single-purpose clearing house business to distribute the tolls, collected by the origin or destination carrier, among all the rail providers on the route. It's the same process in banking --- you can deposit a check drawn on any bank, and a clearing house process routes the funds. It's the same in public stock trading, you can use any brokerage, but almost all shares are owned by Cede & Co on the corporate books.

You can split out payment and address retention easily: paypal and amazon both offer 3rd party checkout systems. If we're trustbusting Amazon, force that to be an independent company, offered on FRAND terms to all.

I think you could probably seperate into three business the fulfilment included marketplace consignment store from the merchant fulfilled marketplace consignment store from the sold by amazon store. Although, since two of those share facilities, it might be a little tough. Perhaps, splitting off the store from the inventory makes more sense --- if Amazon is just one of the marketplace vendors and the marketplace is a separate entity, that eliminates some of the unfair practices (require marketplace vendors to share sources, contract sources directly to undercut marketplace vendors).

You could also split off warehouses off into individual or regional baby-bell style corporations; vendors could send to whichever warehouses they like, the marketplace could choose which warehouse to use to fulfill orders, each warehouse could be reasonably independent.

Shipping/delivery could clearly be made independent. AWS, too.

I think the point of trustbusting is not necessarily to eliminate monopoly or dominant businesses, it's to reduce the scope of the monopoly such that dominance in one area doesn't become dominance in more areas. If there's a single purpose marketplace corporation, it may be dominant, but if it's prevented from controlling warehousing, shipping, and inventory, there's more room in those areas. After the breakup, you can still come back and do more to address the smaller areas, if they prove to be problematic under the new system.


I have come to a conclusion lately that there is no free lunch. You either optimize and increase risks (of system becoming monopoly like rail roads in the past, or busting like mortgage companies in 2007-8, government becoming tyranny, something like that) or you pay for inefficiencies. I think the free market is really about finding a balance between optimizations and risks. If the company is too optimized and carries significant risk - eventually, it will bust. If the marker is not efficient because of transactions volume - a larger company will inevitably emerge (and if it continues to grow - eventually bust)

EDIT: what really worries people about monopolies is that it takes more than a generation to see these companies bust. And to, again, optimize it, we come up with antitrust laws... But any optimization, in my opinion, costs risks and never is free (as in lunch).


>> Amazon online shopping has a "monopoly" almost purely because the monopoliness makes it much easier and lower friction than trying to find an online store that is 1) somewhat trustworthy 2) consistent UI 3) ships in a reasonable time 4) isn't going to disappear (and importantly, already has my shipping and payment information). I regularly buy stuff online from elsewhere, but I use Amazon more often because it reduces the cognitive load by centralizing/standardizing everything.

This is a firm being successful, not a monopolist. This is exactly what antitrust law shouldn't (and does not) punish!

~antitrust economist


I'd press your claim of efficiency. As we are learning now, efficiency (in the JIT sense of the word) and robustness are a single sliding scale. Our global supply chain was a wonder of human achievement until it hit a snag, and now it's a complete mess.

Using Amazon as an example - the larger they've grown, they've actually become less trustworthy and a worse experience for consumers. Amazon has been stricken with all the negative traits that American culture associates with "big" government - bad user experience, bloat, fraud, and arbitrary bureaucracy. I've naturally found myself wanting to use them less and less, the larger they've gotten.


What’s interesting about Amazon is they don’t seem to have been any worse affected from supply chain issues than anyone else. At least not that I have noticed anecdotally.


The general idea would be breaking things into viable businesses and imposing compatibility with competitors. So, for Amazon, spin off the delivery and allow others to do the driving around, provided they can meet the same requirements around picking up from the warehouse and so on. For Google, spin off the DC operations. And so on.

It becomes harder conceptually if you try to break things in the software layer. Working in Google, I just cannot fathom what would it take to actually allow competing ads platforms in search.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: