Here are some possibilities (not claiming these are practical or a solution):
The first is splitting Amazon-retail from their other products (Alexa, Ring, AWS, etc.)
The second is splitting Amazon-logistics out. I'm not exactly sure how this would work, but it could be argued that Amazon logistics could operate standalone.
FBA as a separate business would be interesting for sure. It would separate out the value of Amazon marketplace from the value of outsourcing your order fulfillment logistics.
All I ever wanted was to be able to search "shipped and sold by Amazon" results only. If this is the heavy hammer needed to get to that end result, so be it.
The first one seems obtuse. If here were some 1960s-style comglomerate that sold shower curtains and offered professional bowling ball waxing, with a 95% monopoly in both...
Why would splitting it out into two monopolies improve anything? Even if there are some unknown, misunderstood synergies between the two businesses, will they not just collude in that way that bridge players do, without any overt actions that could be punished?
>will they not just collude in that way that bridge players do, without any overt actions that could be punished?
Currently, they get that for free by virtue of being the same company. If they were separate companies, it would require actual collusion and may, depending on the specifics, be illegal. It's also be less likely to occur since they'd need some motivation to do so. Why would the shower curtain company even attempt to collude with the bowling ball waxing business?
> If they were separate companies, it would require actual collusion a
I don't play bridge, never learned. A game consists of two teams of two players each. My vague understanding is that you might win if a partner gets ahold of a card that you discard, but that you can't know what card that is without saying so (cheating).
Good players can tell what cards their partners might need, and discard anyway. This isn't cheating, it's just good play.
In a court of law, the former is definitely some violation of antitrust law. But the latter is just good business. And even if it weren't, no overt acts have been committed that could lead to successful prosecutorial outcomes.
Philosophically, it may still be collusion. Good luck doing anything about it.
> It's also be less likely to occur since they'd need some motivation to do so.
Game theory supplies that just fine. If I have a monopoly in A, and you a monopoly in B, then I might just protect your monopoly in B without you requesting that, without you giving me instructions.
And from that point on, if you see a place where you can hurt or help me, you might choose to help without me requesting it or asking. Because if my monopoly on A is hurt... I can no longer afford to help you without you asking. Supporting me is a no-brainer. At least if the other party isn't a complete imbecile, they can see it. While there's no accounting for stupidity (second most powerful force after compound interest), we can already assume that the leadership of these companies is non-stupid just because of the success they've already achieved.
> Why would the shower curtain company even attempt to collude with the bowling ball waxing business?
When they break these companies up, they don't fire 100% of management and rehire. Most of them know each other. Most are friendly with each other, if not friends. Sure, employee attrition will eventually break those bonds ten years down the road. But they already know how each other thinks.
Silent collusion, without so much as a wink or a nod, is definitely within the realm of possibility. The previously existing social connections are present. Their goals were largely aligned before, and nothing has occurred to de-align them. Mutual success may be easier than individual success. A sort of "corporate altruism" can emerge.
Evolution does this shit all the time with symbiotes. It's not like their colluding to co-evolve with each other. You won't find conspiracy evidence in the fossil record that they just started helping each other, it only proved to be a better strategy than individuality, and so they went with it.
To a vastly lesser extent than when it's one company with one board and one CEO in charge of multiple business units that would be unrelated after the breakup. If you break off Costco's chicken selling business from the rest of the store, the chicken business isn't going to keep their prices so low to boost the other company's sales because it costs them too much money to do so and they get far less benefit from boosting the other company's sales than when they were the same company.
> If you break off Costco's chicken selling business from the rest of the store,
It sure seems like that, but in the real world we do have many counter-examples where a singular company does do the loss leader thing for a long while, going into the red year after year. Often, it's never clear why or how they could think this was a winning business model.
The chicken-selling business might do the same, and their justifications for doing that might never be sound, they might not even be the reasons why they're doing it, but just some post hoc rationalization to themselves and the real reason is poorly understood by all involved.
If the rest of Costco sees this happening, are they going to step in and sabotage the chicken-sellers? Lord knows it drives some traffic to their stores (they might be unwilling to study that and write reports, as it could then be used against them if the FTC comes after them again, but they'd still notice). If they can subtly act to somehow ease the pressure, or make that worthwhile to the chicken-sellers without leaving a paper trail, they'd be fools not to do it right?
Of course, without exotic economic principles, that can't last forever. But unless you're claiming that there are business units for AWS are deep into loss-leading, your analogy only applies very generally and it's difficult to see how they could fall into similar snags.
The latter seems plausible to me, but the former makes no sense. If the problem is that Amazon-retail has too much market power, how would splitting off AWS help with that problem? Amazon-retail would still have the same fraction of the retail market as before.
Splitting off AWS wouldn't be enough on its own, but I think the argument would be that Amazon makes a substantial amount of its operating income from AWS (70% in the latest quarter: https://ir.aboutamazon.com/news-release/news-release-details...), which allows it to fund anti-competitive marketplace practices like undercutting competitors at a loss to force them out of business. It would be more difficult to sustain such practices without AWS income.
Amazon marketplace, AWS, Whole Foods, Twitch, Fulfillment by Amazon, and maybe Amazon Studios could all be separate companies. I think Amazon marketplace would also be forced to end some practices like not allowing sellers to have lower prices elsewhere.
If you imagine the situation in reverse, where they were two independent companies it would make very little sense to merge them. One is a high margin technology business and the other a low margin distribution company. There would be very little synergy.
He thinks Amazon is acting in good faith, and he's also an ideologue who doesn't see anything wrong with Amazon deplatforming anyone who passes along lower costs on other platforms.
Amazon Retail (Vendor Central), Amazon Marketplace (Seller Central), and Amazon Brands (Amazon Basics, Pinzon, etc.) probably need to be separate companies. You also have decide where Amazon Ads and Amazon Registry will live.
Currently these four services all exist under the same roof and it gives Amazon basically unlimited power to unilaterally destroy entire companies or brands with zero consequences and no effort.
One option is to divide AMZ along business units. Thats the natural, easy way to do so. Itd cause least loss of value to the shareholders.
I disagree. AMZ babies would still be essentially monopolistic in their markets. Nor do I think shareholder value, after being inflated through monopolistic abuse, is sacrisanct.
Divide AMZ slicing through the business units. Make a half dozen vertically integrated baby AMZs. They all have access to AMZ IP, held by a separate company that can license it out to third parties (ie if they don't license their web tech to, say, Best Buy this AMZ baby has no revenue and goes to receivership since all other AMZ babies don't have to pay licensing fees).
EDIT: I almost forgot liability. Transfer all current known and unknown AMZ liabilities, including fraud and sale of fake goods, to the baby AMZs.
I’m suspicious, I think most people underestimate how hard it would be to split companies.
If you have everyone in the company go dig holes for 24 months, the stock price would tank and much value would be destroyed.
Likewise splitting is a massive effort that amounts to burning resources that wouldn’t otherwise be spent.
Now part of the point is to destroy the monopoly, which reduces company value too. This is a subtle point; we are probably ok with this because implicitly this is profit extracted as rent and which should really go to competitors. We’d probably say the value here is being redistributed.
But the monopoly rent is different from the actual OpEx and CapEx required to restructure. I really want to see even a rough quantitative analysis here before making any commitments.
The approach you put forth here seems maximally complex and inefficient as an end state, FWIW. Shared access is going to be a nightmare to operationalize. Much better to cut at existing business unit org boundaries.
What does that look like? What would Amazon look like if broken up?
In the end we all know lobbyists from Amazon will somehow influence their outcome.