15 ms·

Laws of Tech Economics: Commoditize Your Complement

403 points by gwern 9y ago

hemantv 9y ago
I guess same with Amazon they are trying to make brands commodity so that they can extract maximum value out of their supply chain.
- jonwachob91 9y ago
  I'd argue that Amazon commoditized shipping... That's really where they made their name and entrenched their brand recognition.
  - ams6110 9y ago
    Shipping, catalog production, and order processing. That sounds about right.
  - gowld 9y ago
    That's not what "commoditized" means. You don't make your name and brand with a commodity. "commoditize" means making lots of suppliers so there is no producer profit in the market for the commodity. When you shop at Amazon, you don't buy shipping from whoever's cheapest, Amazon provides the shipping via a complex logistics system interfacing with external partners.
- petra 9y ago
  It's true, but Amazon also has a strong push towards it's own brands, and they also move towards fully integrated logistics, so maybe their ultimate strategy is of vertical integration and it was so all along ?
hinkley 9y ago
This bums me out. I always liked the idea of two companies in complementary spaces working together toward something and maybe merging some day far in the future. Instead we’re trying to make each other trivial.
- LeifCarrotson 9y ago
  Why just two companies? The article describes one company with a chokehold in one spot of the vertical and many other companies producing commodity complements. But there are lots of verticals near each other, so there can be many people at one level working with commodities at another level, who likewise believe that the people above/below them are also generating a commodity complement. What we end up with instead is a global economy of people generating value and building things that make lives better, which I think is awesome! Generating value is the most legitimate reason to have people pay you - you don't have to extract rent from a monopoly position to be a viable business.
- ajross 9y ago
  > Instead we’re trying to make each other trivial. Which is about as good a definition of "technological progress" as you are going to find. I for one am happy at how economically trivial housing, transport and food production have become over the last thousand years.
- timerol 9y ago
  This doesn't bum me out at all. Looking at the headline examples given, they all seem to move tech in a better direction. If more of the trivial things are commoditized, then more effort goes into competing on the things that matter. Example headlines FTA: IBM Spends Millions to Develop Open Source Software; Netscape Open Sources Their Web Browser; Transmeta Hires Linus, Pays Him To Hack on Linux
- draw_down 9y ago
  Welcome to capitalism!
shmerl 9y ago
> Vive and Oculus have incentives to cooperate… for now. But there’s constant pressure for a betrayal when a player gets desperate or decides the market has matured and it’s time to break for the finish line OpenXR seems to be moving forward. This will hopefully solve this VR lock-in mess.
ucaetano 9y ago
Another way that I like to express that is "create a desert of profitability around you". I once had a strategy professor define the Google business model somewhat like that, where "Google tries to make every other business around it free or irrelevant". It results in a few different effects: - By reducing the cost of other links of the value chain, there is more money available to spend on the links you actually generate revenue on. This shifts profits along the value chain to that link. One example is dramatically reducing the cost of phones and internet access, thus allowing customers to spend more time and money online, which generates revenue and profits for Google - By making the other links free or irrelevant, you reduce the odds that a competitor in those links will strengthen their position and will extract more profits from the rest of the value chain. One example is using Android to prevent a monopoly on the smartphone side. If Apple had a monopoly or near monopoly, it would be able to extract larger economic profits from the other links on the value chain, including Google A desert of profitability shifts consumers to you, and keeps competitors away.
- sah2ed 9y ago
  FWIW, the expression "create a desert of profitability around you" is more popularly expressed as "create an economic moat".
  - Symmetry 9y ago
    The ideas are more different than the analogies are. A moat would be things like patents, data, expertise, etc that a competitor would need to get to match you and which are hard to develop. That doesn't have anything to do with the profitability of nearby market niches.
    - ucaetano 9y ago
      Exactly! A moat is less efficient than a desert. A moat is highly susceptible to niche entrants who might expand into your main market after, but a desert leaves no to little money on the table for anyone else.
peterarmstrong 9y ago
This is a good point in general and about the PC in particular, but it is totally wrong about Netscape. The complement of a web browser isn't the web servers, it's the OS--and vice versa. So, Netscape tries to commoditize the OS, and MS commoditized and destroyed the browser market, and then nearly themselves with antitrust. Also, Netscape was initially free in the early days, but had become a product which cost money until IE came out and was free. The plan was not "free browser". (Read Ben Horowitz's book...)
- sah2ed 9y ago
  Several things can be complements, it just depends on how you perceive them. In Netscape's case, their strategy shifted from selling web browsers to consumers in the hope of commoditizing the OS [0] to selling more web servers to businesses, after Microsoft made it impossible for them to compete[1] with IE which was offered for free. [0] https://en.wikipedia.org/wiki/Netscape#History_of_the_Netscape_Communications_Corporation https://en.wikipedia.org/wiki/Netscape#History_of_the_Netsca... [1] https://www.cnet.com/news/netscape-cuts-prices-on-retail-products/ https://www.cnet.com/news/netscape-cuts-prices-on-retail-pro...
saosebastiao 9y ago
This is just tech rebranding of a really old concept. There are mountains of evidence of this strategy being employed back as far as the industrial revolution. I don't know why, but this sort of crap really brings out the eye-rolls. Maybe it's the fact that this information is widely and freely available, but NIH syndrome makes people feel like they've discovered some new and novel concept about the world. And then you get big-ego personalities who spend massive amounts of time congratulating each other about it. Like when Peter Thiel starting talking about his novel strategy of explicitly trying to not compete with others and aiming for a monopoly. You can find that explicitly described in complete detail in any ordinary marketing textbook in the chapter titled Market Segmentation, Targeting, and Positioning. But for months people couldn't stop talking about how Thiel was a genius for this new idea. I get the whole concept of XKCD's lucky 10,000. This is different. This would be like taking credit for inventing mentos and coke bombs just because someone hasn't heard of it before.
- emiliobumachar 9y ago
  Okay, what's the name of the old concept that's the same as "Commoditize Your Complement"? The textbook chapter title, if you will?
  - saosebastiao 9y ago
    It's common and mundane enough to not even have a name (at least up until the point where it was "branded"). In a strategy textbook, you would find the ideas discussed as a bulletpoint tactic to use for two sided markets or format wars. It is often mentioned as a strategy for aftermarket parts and services in a growth market. In pricing, it's a minor variation of the razor-and-blades model where you outsource or standardize the "razor".
    - scarface74 9y ago
      That's not the same thing. The razor and blades model is where you are both selling the razor at below market prices and the blades at a markup. "Commoditing your complements" is about you making the razor blades interchangeable where multiple manufacturers are competing to sell razor blades at low margins while you make all of the money selling razors. Apple commoditized apps and the operating system by giving them away for free, as did Google. Microsoft commoditized PC manufacturers, etc.
sah2ed 9y ago
This article illustrates why I agree with criticisms [0] of Michael Porter's five forces framework as being incomplete without the sixth force -- threat of complementors seeking to commoditize an industry. [0] https://en.wikipedia.org/wiki/Porter%27s_five_forces_analysis#Criticisms https://en.wikipedia.org/wiki/Porter%27s_five_forces_analysi...
strong_silent_t 9y ago
I like the book Information Rules by Shapiro and Varian as a review of these kinds of concepts in general. I don't think it is necessary to even view these topics as strategy, to a certain extent it is just the nature of the incentives and there will be a tendency for things to work out along these lines. Here is a snapshot of the table of contents, from this it should be pretty clear if it is interesting to you: https://imgur.com/a/2yU6OpA https://imgur.com/a/2yU6OpA
- davidw 9y ago
  I would also highly recommend the same book. "Technology changes, economic laws do not" https://amzn.to/2I9NTzp https://amzn.to/2I9NTzp Among other things, Hal Varian went on to work at Google as an economist.
- sah2ed 9y ago
  Agreed, the book does a solid review of the underlying principles. They gave an example of how Intel used their AGP standard as part of their commoditize your complement tactic except they characterized it as maximizing the value of your technology (p. 197-198): "In choosing between openness and control, remember that your ultimate goal is to maximize the value of your technology, not your control over it."
RcouF1uZ4gsC 9y ago
One of the things the tech economies try to commedituze is developers. While you sounding nice, the whole tech company push for "everyone can code" is an effort to commodetize developers. Having free and open source compilers and languages also serves to reduce the cost of entry and commodetize developers. In the past, developers would pay hundreds of dollars for compilers and languages (which is actually cheap compared to what other professions pay for their tools). Developers volunteer to teach new people to code. Developers spend their free time making open source software that is free for companies to use. Developers contribute to free languages and tools. Finally, developers complain that they are taken advantage by business people who think they are stupid! Software developers are the only professionals which actively work to commodetize themselves.
- HillaryBriss 9y ago
  my sense is that a lot of developers create or contribute to FOSS projects in order to hone their skills and promote their own personal name (brand). they're trying to gain traction in the slippery, highly competitive landscape. of course, they're also trying to make a contribution to the larger ecosystem, because they feel they've benefited from the contributions of others.
imhoguy 9y ago
I wonder is this even possible in small scale. Does anybody have some intresting examples of startups or even side-projects "commoditizing their complement"?
aaavl2821 9y ago
Not unlike what YC is doing with startup school and angel investor school -- increase number of startups, increase number of seed investors, increase your brand in the mind of both, and have a monopoly on the pre-seed stage
aaavl2821 9y ago
Hospitals and integrated health systems sort of do this, though in a convoluted way. They buy up other hospitals and outpatient clinics and use scale to negotiate higher rates with insurers. Insurers get squeezed, and have to cut payments to less powerful independent physicians / hospitals. This often forces those independent providers to sell to the big hospital systems. It's basically impossible to start / run an independent primary care practice these days in some geographies. There's only one independent PCP in Oakland these days due to dominance of sutter So not the same means of making complements unprofitable as the tech examples, but the same end -- the hospitals and some hospital owned outpatient clinics drive all the profits, everything else is like a loss leader that prevents competition from springing up around the hospital
- HillaryBriss 9y ago
  In some states, a distinctive aspect of the hospital provisioning landscape is "Certificate of Need" (CON) laws. Since new hospitals cannot be constructed without proving a "need", the certificate-of-need system grants monopoly privileges to already existing hospitals. [At least one lawmaker has] argued that the true motivation behind certificate-of-need legislation is that "large hospitals are... trying to make money by eliminating competition" under the pretext of using monopoly profits to provide better patient care. More at https://en.wikipedia.org/wiki/Certificate_of_need https://en.wikipedia.org/wiki/Certificate_of_need
  - aaavl2821 9y ago
    I think many health systems are using the ACA as a pretext for establishing further monopoly power. Basically the line for the public is "we need all providers in one system so we have all the data and so we can refer patients to the right site of care" In reality they are bringing all providers under their umbrella to control patients i.e. Market share, and to drive patients to the site of care that is most profitable to them under the constraints of minimizing malpractice risk and quality fines. ACOs are in many cases just a way to increase patient volume at a discount to payers and have nothing to do with risk sharing Hospitals are an incredibly political powerful entity. The AHA (hospital industry lobbying group) spends about as much as Phrma (pharma lobbying group), but hospitals have massive grassroots political support bc they are huge employers whereas the public hates pharma /endrant
jackconnor 9y ago
Joel Spolsky continues his streak as most-influential tech writer of all time. Great article.
ksec 9y ago
How is this different from devaluing / commoditizing your competitor's product or services? Because it isn't just tech, and it isn't / doesn't have to be complement. Any company that has a cash cow and fail to find a new way for growth could use that cash to commoditizing its competitor product or services, this happens in every industry. Jeff Bezos, possibly the king of commoditization, "Your margin is my opportunity."
fhrow4484 9y ago
Interesting article, I'm trying to figure out: What are people trying to commoditize with Kubernetes. But even more importantly: How is this helping them? (since in my head the 'what' being commoditized is the same thing the big players contributing to Kubernetes make profits with)
- chubot 9y ago
  Google and everyone supporting Kubernetes are trying to prevent AWS from becoming the de facto cloud "system call API", i.e. for provisioning computing resources like CPU/memory/storage/network, and for higher level services like monitoring, access control, backup, dev tools, etc. They want to make it easier to move off of the market leader, AWS. If a customer's programs are written to AWS APIs, it becomes incredibly expensive to even try a new cloud provider. So basically every other cloud provider besides AWS wants standard interfaces. They can all rent more computing resources if it's easier for customers to move off of AWS. That's why Microsoft supports Kubernetes. And that's why everyone got on board with Docker too -- it's (supposedly) a standard application format, or at least a cloud-vendor-neutral one. It's exactly analogous to Windows vs. Unix. Windows locked users into its proprietary APIs, with exactly one implementation, and made tremendous profits. With POSIX, you can port applications from Solaris to IBM Unix to Linux, etc. Kubernetes was called "POSIX for the cloud" as far as I remember. It's related to the "commoditize your complements" model, but I don't think that's exactly the right way to describe it. Maybe it's more like "commoditizing your competitor's product", to erode their profits. In theory I don't see a problem with AWS owning 95% of the cloud market share and Google still making billions of dollars on ads (with free complements like search, apps, Android, and the browser). I think AWS is more threatening to Microsoft, because Microsoft sells enterprise products like SQL Server and so forth. Cloud services are the complement to a lot of enterprise products that MS sells. Google isn't nearly as big an enterprise business as Microsoft is, and the customers for AWS/Cloud are typically enterprises and not "consumers".
chubot 9y ago
Classic example from Google: spending $106 million to release free video codecs in 2009. Video codecs are a complement of YouTube, and MP4 licensing fees were apparently enough of a barrier to justify this purchase. https://searchengineland.com/google-buys-video-infrastructure-company-on2-23628 https://searchengineland.com/google-buys-video-infrastructur... With 9 years of hindsight, how did it turn out? I'm honestly not sure, as I haven't followed the state of video codecs. It seems like a lot of video is viewed on Apple devices, which only have hardware support for MP4?
ryanackley 9y ago
Just realized that Atlassian does this. Bitbucket commoditizes code hosting and HipChat/Stride commoditizes team chat. Both are loss leaders meant to protect their core business: Jira
javajosh 9y ago
Can it be used to explain why Google discontinued certain products, especially Reader?
hi41 9y ago
I read this many years back at joelonsoftware. Joel is such an amazing thinker. His ideas are so original. I don't know how one thinks such original thoughts. I can follow orders but thinking for myself is very hard.

403 points by gwern 9y ago

I guess same with Amazon they are trying to make brands commodity so that they can extract maximum value out of their supply chain.

jonwachob91 9y ago
I'd argue that Amazon commoditized shipping... That's really where they made their name and entrenched their brand recognition.
- ams6110 9y ago
  Shipping, catalog production, and order processing. That sounds about right.
- gowld 9y ago
  That's not what "commoditized" means. You don't make your name and brand with a commodity. "commoditize" means making lots of suppliers so there is no producer profit in the market for the commodity. When you shop at Amazon, you don't buy shipping from whoever's cheapest, Amazon provides the shipping via a complex logistics system interfacing with external partners.
petra 9y ago
It's true, but Amazon also has a strong push towards it's own brands, and they also move towards fully integrated logistics, so maybe their ultimate strategy is of vertical integration and it was so all along ?

hinkley 9y ago

This bums me out. I always liked the idea of two companies in complementary spaces working together toward something and maybe merging some day far in the future. Instead we’re trying to make each other trivial.

LeifCarrotson 9y ago
Why just two companies? The article describes one company with a chokehold in one spot of the vertical and many other companies producing commodity complements. But there are lots of verticals near each other, so there can be many people at one level working with commodities at another level, who likewise believe that the people above/below them are also generating a commodity complement. What we end up with instead is a global economy of people generating value and building things that make lives better, which I think is awesome! Generating value is the most legitimate reason to have people pay you - you don't have to extract rent from a monopoly position to be a viable business.
ajross 9y ago
> Instead we’re trying to make each other trivial. Which is about as good a definition of "technological progress" as you are going to find. I for one am happy at how economically trivial housing, transport and food production have become over the last thousand years.
timerol 9y ago
This doesn't bum me out at all. Looking at the headline examples given, they all seem to move tech in a better direction. If more of the trivial things are commoditized, then more effort goes into competing on the things that matter. Example headlines FTA: IBM Spends Millions to Develop Open Source Software; Netscape Open Sources Their Web Browser; Transmeta Hires Linus, Pays Him To Hack on Linux
draw_down 9y ago
Welcome to capitalism!

shmerl 9y ago

> Vive and Oculus have incentives to cooperate… for now. But there’s constant pressure for a betrayal when a player gets desperate or decides the market has matured and it’s time to break for the finish line OpenXR seems to be moving forward. This will hopefully solve this VR lock-in mess.

ucaetano 9y ago

Another way that I like to express that is "create a desert of profitability around you". I once had a strategy professor define the Google business model somewhat like that, where "Google tries to make every other business around it free or irrelevant". It results in a few different effects: - By reducing the cost of other links of the value chain, there is more money available to spend on the links you actually generate revenue on. This shifts profits along the value chain to that link. One example is dramatically reducing the cost of phones and internet access, thus allowing customers to spend more time and money online, which generates revenue and profits for Google - By making the other links free or irrelevant, you reduce the odds that a competitor in those links will strengthen their position and will extract more profits from the rest of the value chain. One example is using Android to prevent a monopoly on the smartphone side. If Apple had a monopoly or near monopoly, it would be able to extract larger economic profits from the other links on the value chain, including Google A desert of profitability shifts consumers to you, and keeps competitors away.

sah2ed 9y ago
FWIW, the expression "create a desert of profitability around you" is more popularly expressed as "create an economic moat".
- Symmetry 9y ago
  The ideas are more different than the analogies are. A moat would be things like patents, data, expertise, etc that a competitor would need to get to match you and which are hard to develop. That doesn't have anything to do with the profitability of nearby market niches.
  - ucaetano 9y ago
    Exactly! A moat is less efficient than a desert. A moat is highly susceptible to niche entrants who might expand into your main market after, but a desert leaves no to little money on the table for anyone else.

peterarmstrong 9y ago

This is a good point in general and about the PC in particular, but it is totally wrong about Netscape. The complement of a web browser isn't the web servers, it's the OS--and vice versa. So, Netscape tries to commoditize the OS, and MS commoditized and destroyed the browser market, and then nearly themselves with antitrust. Also, Netscape was initially free in the early days, but had become a product which cost money until IE came out and was free. The plan was not "free browser". (Read Ben Horowitz's book...)

sah2ed 9y ago
Several things can be complements, it just depends on how you perceive them. In Netscape's case, their strategy shifted from selling web browsers to consumers in the hope of commoditizing the OS [0] to selling more web servers to businesses, after Microsoft made it impossible for them to compete[1] with IE which was offered for free. [0] https://en.wikipedia.org/wiki/Netscape#History_of_the_Netscape_Communications_Corporation https://en.wikipedia.org/wiki/Netscape#History_of_the_Netsca... [1] https://www.cnet.com/news/netscape-cuts-prices-on-retail-products/ https://www.cnet.com/news/netscape-cuts-prices-on-retail-pro...

saosebastiao 9y ago

This is just tech rebranding of a really old concept. There are mountains of evidence of this strategy being employed back as far as the industrial revolution. I don't know why, but this sort of crap really brings out the eye-rolls. Maybe it's the fact that this information is widely and freely available, but NIH syndrome makes people feel like they've discovered some new and novel concept about the world. And then you get big-ego personalities who spend massive amounts of time congratulating each other about it. Like when Peter Thiel starting talking about his novel strategy of explicitly trying to not compete with others and aiming for a monopoly. You can find that explicitly described in complete detail in any ordinary marketing textbook in the chapter titled Market Segmentation, Targeting, and Positioning. But for months people couldn't stop talking about how Thiel was a genius for this new idea. I get the whole concept of XKCD's lucky 10,000. This is different. This would be like taking credit for inventing mentos and coke bombs just because someone hasn't heard of it before.

emiliobumachar 9y ago
Okay, what's the name of the old concept that's the same as "Commoditize Your Complement"? The textbook chapter title, if you will?
- saosebastiao 9y ago
  It's common and mundane enough to not even have a name (at least up until the point where it was "branded"). In a strategy textbook, you would find the ideas discussed as a bulletpoint tactic to use for two sided markets or format wars. It is often mentioned as a strategy for aftermarket parts and services in a growth market. In pricing, it's a minor variation of the razor-and-blades model where you outsource or standardize the "razor".
  - scarface74 9y ago
    That's not the same thing. The razor and blades model is where you are both selling the razor at below market prices and the blades at a markup. "Commoditing your complements" is about you making the razor blades interchangeable where multiple manufacturers are competing to sell razor blades at low margins while you make all of the money selling razors. Apple commoditized apps and the operating system by giving them away for free, as did Google. Microsoft commoditized PC manufacturers, etc.

sah2ed 9y ago

This article illustrates why I agree with criticisms [0] of Michael Porter's five forces framework as being incomplete without the sixth force -- threat of complementors seeking to commoditize an industry. [0] https://en.wikipedia.org/wiki/Porter%27s_five_forces_analysis#Criticisms https://en.wikipedia.org/wiki/Porter%27s_five_forces_analysi...

strong_silent_t 9y ago

I like the book Information Rules by Shapiro and Varian as a review of these kinds of concepts in general. I don't think it is necessary to even view these topics as strategy, to a certain extent it is just the nature of the incentives and there will be a tendency for things to work out along these lines. Here is a snapshot of the table of contents, from this it should be pretty clear if it is interesting to you: https://imgur.com/a/2yU6OpA https://imgur.com/a/2yU6OpA

davidw 9y ago
I would also highly recommend the same book. "Technology changes, economic laws do not" https://amzn.to/2I9NTzp https://amzn.to/2I9NTzp Among other things, Hal Varian went on to work at Google as an economist.
sah2ed 9y ago
Agreed, the book does a solid review of the underlying principles. They gave an example of how Intel used their AGP standard as part of their commoditize your complement tactic except they characterized it as maximizing the value of your technology (p. 197-198): "In choosing between openness and control, remember that your ultimate goal is to maximize the value of your technology, not your control over it."

RcouF1uZ4gsC 9y ago

One of the things the tech economies try to commedituze is developers. While you sounding nice, the whole tech company push for "everyone can code" is an effort to commodetize developers. Having free and open source compilers and languages also serves to reduce the cost of entry and commodetize developers. In the past, developers would pay hundreds of dollars for compilers and languages (which is actually cheap compared to what other professions pay for their tools). Developers volunteer to teach new people to code. Developers spend their free time making open source software that is free for companies to use. Developers contribute to free languages and tools. Finally, developers complain that they are taken advantage by business people who think they are stupid! Software developers are the only professionals which actively work to commodetize themselves.

HillaryBriss 9y ago
my sense is that a lot of developers create or contribute to FOSS projects in order to hone their skills and promote their own personal name (brand). they're trying to gain traction in the slippery, highly competitive landscape. of course, they're also trying to make a contribution to the larger ecosystem, because they feel they've benefited from the contributions of others.

imhoguy 9y ago

I wonder is this even possible in small scale. Does anybody have some intresting examples of startups or even side-projects "commoditizing their complement"?

aaavl2821 9y ago

Not unlike what YC is doing with startup school and angel investor school -- increase number of startups, increase number of seed investors, increase your brand in the mind of both, and have a monopoly on the pre-seed stage

aaavl2821 9y ago

Hospitals and integrated health systems sort of do this, though in a convoluted way. They buy up other hospitals and outpatient clinics and use scale to negotiate higher rates with insurers. Insurers get squeezed, and have to cut payments to less powerful independent physicians / hospitals. This often forces those independent providers to sell to the big hospital systems. It's basically impossible to start / run an independent primary care practice these days in some geographies. There's only one independent PCP in Oakland these days due to dominance of sutter So not the same means of making complements unprofitable as the tech examples, but the same end -- the hospitals and some hospital owned outpatient clinics drive all the profits, everything else is like a loss leader that prevents competition from springing up around the hospital

HillaryBriss 9y ago
In some states, a distinctive aspect of the hospital provisioning landscape is "Certificate of Need" (CON) laws. Since new hospitals cannot be constructed without proving a "need", the certificate-of-need system grants monopoly privileges to already existing hospitals. [At least one lawmaker has] argued that the true motivation behind certificate-of-need legislation is that "large hospitals are... trying to make money by eliminating competition" under the pretext of using monopoly profits to provide better patient care. More at https://en.wikipedia.org/wiki/Certificate_of_need https://en.wikipedia.org/wiki/Certificate_of_need
- aaavl2821 9y ago
  I think many health systems are using the ACA as a pretext for establishing further monopoly power. Basically the line for the public is "we need all providers in one system so we have all the data and so we can refer patients to the right site of care" In reality they are bringing all providers under their umbrella to control patients i.e. Market share, and to drive patients to the site of care that is most profitable to them under the constraints of minimizing malpractice risk and quality fines. ACOs are in many cases just a way to increase patient volume at a discount to payers and have nothing to do with risk sharing Hospitals are an incredibly political powerful entity. The AHA (hospital industry lobbying group) spends about as much as Phrma (pharma lobbying group), but hospitals have massive grassroots political support bc they are huge employers whereas the public hates pharma /endrant

jackconnor 9y ago

Joel Spolsky continues his streak as most-influential tech writer of all time. Great article.

ksec 9y ago

How is this different from devaluing / commoditizing your competitor's product or services? Because it isn't just tech, and it isn't / doesn't have to be complement. Any company that has a cash cow and fail to find a new way for growth could use that cash to commoditizing its competitor product or services, this happens in every industry. Jeff Bezos, possibly the king of commoditization, "Your margin is my opportunity."

fhrow4484 9y ago

Interesting article, I'm trying to figure out: What are people trying to commoditize with Kubernetes. But even more importantly: How is this helping them? (since in my head the 'what' being commoditized is the same thing the big players contributing to Kubernetes make profits with)

chubot 9y ago
Google and everyone supporting Kubernetes are trying to prevent AWS from becoming the de facto cloud "system call API", i.e. for provisioning computing resources like CPU/memory/storage/network, and for higher level services like monitoring, access control, backup, dev tools, etc. They want to make it easier to move off of the market leader, AWS. If a customer's programs are written to AWS APIs, it becomes incredibly expensive to even try a new cloud provider. So basically every other cloud provider besides AWS wants standard interfaces. They can all rent more computing resources if it's easier for customers to move off of AWS. That's why Microsoft supports Kubernetes. And that's why everyone got on board with Docker too -- it's (supposedly) a standard application format, or at least a cloud-vendor-neutral one. It's exactly analogous to Windows vs. Unix. Windows locked users into its proprietary APIs, with exactly one implementation, and made tremendous profits. With POSIX, you can port applications from Solaris to IBM Unix to Linux, etc. Kubernetes was called "POSIX for the cloud" as far as I remember. It's related to the "commoditize your complements" model, but I don't think that's exactly the right way to describe it. Maybe it's more like "commoditizing your competitor's product", to erode their profits. In theory I don't see a problem with AWS owning 95% of the cloud market share and Google still making billions of dollars on ads (with free complements like search, apps, Android, and the browser). I think AWS is more threatening to Microsoft, because Microsoft sells enterprise products like SQL Server and so forth. Cloud services are the complement to a lot of enterprise products that MS sells. Google isn't nearly as big an enterprise business as Microsoft is, and the customers for AWS/Cloud are typically enterprises and not "consumers".

chubot 9y ago

Classic example from Google: spending $106 million to release free video codecs in 2009. Video codecs are a complement of YouTube, and MP4 licensing fees were apparently enough of a barrier to justify this purchase. https://searchengineland.com/google-buys-video-infrastructure-company-on2-23628 https://searchengineland.com/google-buys-video-infrastructur... With 9 years of hindsight, how did it turn out? I'm honestly not sure, as I haven't followed the state of video codecs. It seems like a lot of video is viewed on Apple devices, which only have hardware support for MP4?

ryanackley 9y ago

Just realized that Atlassian does this. Bitbucket commoditizes code hosting and HipChat/Stride commoditizes team chat. Both are loss leaders meant to protect their core business: Jira

javajosh 9y ago

Can it be used to explain why Google discontinued certain products, especially Reader?

hi41 9y ago

I read this many years back at joelonsoftware. Joel is such an amazing thinker. His ideas are so original. I don't know how one thinks such original thoughts. I can follow orders but thinking for myself is very hard.