To be fair, Samsung's divisions having guns pointed at each other is nothing new. This is the same conglomerate that makes their own chip division fight for placement in their own phones, constantly flip-flopping between using Samsung or Qualcomm chips at the high end, Samsung or Mediatek chips at the low end, or even a combination of first-party and third-party chips in different variants of ostensibly the same device.
It's a forcing function that ensures the middle layers of a vertically integrated stack remain market competitive and don't stagnate because they are the default/only option
Sears had horizontal market where all of it did basically the same thing. Samsung is a huge conglomerate of several completely different vertical with lots of redundant components.
It makes absolutely no sense to apply the lessons from one into the other.
Sears had horizontal market where all of it did basically the same thing. Samsung is a huge conglomerate of several completely different vertical with lots of redundant components.
Sears was hardly horizontal. It was also Allstate insurance and Discover credit cards, among other things.
Ok. And if it did divide on the borders of insurance and payment services, the reorganization wouldn't have been complete bullshit and may even have been somewhat successful.
I think what the GP was referring to was the "new" owner of Sears, who reorganized the company into dozens of independent business units in the early 2010s (IT, HR, apparel, electronics, etc). Not departments, either; full-on internal businesses intended as a microcosm of the free market.
Each of these units were then given access to an internal "market" and directed to compete with each other for funding.
The idea was likely to try and improve efficiency... But what ended up happening is siloing increased, BUs started infighting for a dwindling set of resources (beyond normal politics you'd expect at an organization that size; actively trying to fuck each other over), and cohesion decreased.
It's often pointed to as one of the reasons for their decline, and worked out so badly that it's commonly believed their owner (who also owns the company holding their debt and stands to immensely profit if they go bankrupt) desired this outcome... to the point that he got sued a few years ago by investors over the conflict of interest and, let's say "creative" organizational decisions.
This happened at a place where I worked years ago, but not as 'on purpose.' We were a large company where most pieces depended on other pieces, and everything was fine - until a new CEO came in who started holding the numbers of each BU under a microscope. This led to each department trying to bill other departments as an enterprise customer, who then retaliated, which then led to internal departments threatening to go to competitors who charged less for the same service. Kinda stupid how that all works - on paper it would have made a few departments look better if they used a bottom barrel competitor, but in reality the company would have bled millions of dollars as a whole...all because one rather large BU wanted to goose its numbers.
Why is that a bad thing? If an internal department that’s not core to their business is less efficient than an external company - use the external company.
Anecdote: Even before Amazon officially killed Chime, everyone at least on the AWS side was moving to officially supported Slack.
I guess it depends on circumstances, but it boils down to each department only cost others some marginal cost in practice.
Imagine a hosting company and a dns company, both with plenty of customers and capacity. The hosting company says... I'll host your DNS site, if you provide DNS to our hosting site. Drop in the bucket for each.
One year the DNS company decides it needs to show more revenue, so will begin charging the hosting company $1000/yr, and guess what the hosting company says the same. Instead, they each get mad and find $500/yr competitors. What was accomplished here?
Further, it just looks bad in many cases. Imagine if Amazon.com decided AWS was too expensive, and decided to move their stuff off to say, Azure only. That wouldn't be a great look for AWS and in turn hurts...Amazon.
I do get your point, but there are a lot of... intangibles about being in a company together.
There is more politics than you think within Amazon Retail about moving compute over to AWS. I’m not sure how much of Amazon Retail runs on AWS instead of its own infrastructure (CDO).
I know one project from Amazon got killed because their AWS bill was too high. Yeah AWS charges Amazon Retail for compute when they run on AWS hardware.
As a rule, organizations are created to avoid the transaction costs on those detail tasks. If you externalize every single supporting task into a market, you will be slowed down to a drag, won't be able to use most competitive advantages, and will pay way more than doing them in house.
But removing the market competition is a breeding ground for inefficiency. So there's a balance there, and huge conglomerates tying their divisions together serves only to make the competitive ones die by the need to use the services of the inefficient ones.
My four years at AWS kind of indoctrinated me. As they said, everytime you decide to buy vs build, you have to ask yourself “does it make the beer taste better”?
Don’t spend energy on undifferentiated heavy lifting. If you are Dropbox it makes sense to move away from S3 for instance.
to put a finer point on it, it wasn't just competition or rewarding-the-successful, the CEO straight up set them at odds with each other and told them directly to battle it out.
basically "coffee is for closers... and if you don't sell you're fired" as a large scale corporate policy.
That was a bullshit separation of a single horizontal cut of the market (all of those segments did consumer retail sales) without overlap.
The part about no overlaps already made it impossible for them to compete. The only "competition" they had was in the sense of TV gameshow competition where candidates do worthless tasks, judged by some arbitrary rules.
That has absolutely no similarity to how Samsung is organized.
Not sure that the opposite of transfer pricing is nepotism. As far as I know it’s far more common for someone who owns a lake house to assign four weeks a year to each grandkid , than to make them bid real money on it and put that in a maintenance fund or something. Though it’s an interesting idea, it’s not very family friendly
You mean toyota putting bmw engine (supra). Your statement is contradicting as Toyota has TRD, which focuses on the track performance. They just couldn't keep up with the straight six perf+reliability when comparing to their own 2jz
Buying a Supra is stupid. Either buy a proper BMW with the b58/Zf8 speed and get a proper interior or stop being poor and buy an LC500.
Better yet, get a C8 corvette and gap all of the above for a far better value. You can get 20% off msrp on factory orders with C8 corvettes if you know where to look.
They operate with tension. They're supposed to have unified strategic direction from the top, but individual subsidiaries are also expected to be profit centers that compete in the market.
I worked with some supply chain consultants who mentioned "internal suppliers are often worse suppliers than external".
Their point was that service levels are often not as stringently tracked, SLA's become internal money shuffling, but the company as a whole paid the price in lower output/profit. The internal partner being the default allows an amount of complacency, and if you shopped around for a comparable level of service to what's being provided, you can often find it for a better price.
Basically every Galaxy phone comes in two versions. One with Exynos and one with Snapdragon. It's regional though. US always gets the Snapdragon phones while Europe and mostly Asia gets the Exynos version.
My understanding is that the Exynos is inferior in a lot of ways, but also cheaper.
In the past using Snapdragon CPUs for the U.S. made sense due to Qualcomm having much better support for the CDMA frequencies needed by Verizon. Probably no longer relevant since the 5G transition though.
Not one phone, they did this all over the place. Their flagship line did this starting with the Galaxy S7 all the way up to Galaxy S24. Only the most recent Galaxy S25 is Qualcomm Snapdragon only, supposedly because their own Exynos couldn't hit volume production fast enough.
"Galaxy S II" and its aesthetics was already a mere branding shared across at least four different phones with different SoCs, before counting in sub-variants that share same SoCs. This isn't unique to Samsung, nor is it a new phenomenon, just how consumer products are made and sold.
The S23 too was Snapdragon only, allegedly to let the Exynos team catch some breath and come up with something competitive for the following generation. Which they partly did, as the Exynos S24 is almost on par with its Snapdragon brother. A bit worse on photo and gaming performance, a bit better in web browsing, from the benchmarks I remember.
