Oracle's argument is that it shouldn't be delivered as a single contract and should be split among multiple cloud vendors.
Anyone whose used AWS/Azure/Oracle Cloud will tell you why that idea is problematic. Particularly when dealing with a hardened government version of these cloud environments.
The only potential corruption their complaint indicates is the hiring of a decision maker at the DoD. It is a bad look. I'd need to see the timeline/understand more about it to form an opinion.
Thanks! Oracle is an RDB and app provider, mostly web-based.
AWS is a cloud platform provider.
Unless one (or both?) have changed while I wasn't looking... but even then, is the Cloud now so central to Oracle that its existence depends on it?
EDIT thanks for the link. "Oracle data warehouse" isn't Oracle's central business. Also, I'm not a fan of Oracle, but I doubt it's so easy to steal their customers. Finally, how can the loss of one govt contract be so harmful? Few of their customers, not even govt ones, are so security conscious... BTW I realize you're just reporting their claim, not making it.
I don't think the issue is how important the cloud is to Oracle. The threat to Oracle is that AWS provides directly comparable database products (RDS, Redshift, DynamoDB) that don't require the licensing fees and (sometimes) hardware costs that come with Oracle products.
As an aside, I regret not disclosing earlier that I'm an engineer at Amazon. I did not expect to participate in this discussion this much, and also, it just didn't occur to me. My apologies.
Thanks. Oracle already has competing cheaper/free RDBs - though not an integrated suite. (I accept O's app businesses are secondary).
From your first link, compatibility is hard: AWS SCT clearly marks any code that it cannot convert so that you can manually convert it.
I would expect O to be upset that they don't even get a look in to an incredibly lucrative contract, and argue it's bad for the govt's well-being to exclude them. I can't see how or why they'd argue it threatens their existence...
Oh, so you are making the claim! Your comment makes a lot more sense now.
BTW congrats on winning the contract, that's amazing!
I haven't made any claims about Oracle's claims in the filing. Perhaps they are entirely valid. I don't disagree with anything you've said about the details of the competing services. Perhaps you are entirely correct and Oracle has a superior offering.
The thesis of my argument is that, in my opinion, Amazon represents a true threat to Oracle, and that moves like this filing should be examined through that lens.
For the record, my opinions are my own and I have no knowledge of this situation or AWS decision making that is not public.
But you're describing a "disruption": AWS's offering is comparable yet far cheaper. Which sounds like an "existential threat".
Oracle should be immune to this disruption: the same people are doing the same thing to the same data for the same purpose - so, the cloud is a "sustaining" innovation in this context.
Plus, Oracle is a sales company, not an engineering company, and I don't think AWS is oriented to compete that way.
You've softened your thesis to a "threat", and yes I agree of course it is. And yes, their competitive situation will inform their actions.
But... $10 billion dollars is reason enough to file in itself, right? They have huge departments devoted to just this kind of thing. And recall the fuss they made vs Google over Java, for much less (though TBF, that had strategic advantages for them).
I still don't see how this particular contract would have knock-on consequences for Oracle, making it an existential threat. Maybe I'm missing something.
I have not softened my thesis. For the sake of brevity I assumed that just saying threat was sufficient.
> Oracle should be immune to this disruption: the same people are doing the same thing to the same data for the same purpose - so, the cloud is a "sustaining" innovation in this context.
The idea that any company is immune to disruption seems obviously wrong to me. But in this case, it seems very clear. You're correct that people use databases the same way for the same purpose on cloud vs self host -- no one disputes that. What the cloud dramatically alters, among many things, is pay-what-you-use pricing and scalability. AWS database offerings render license fees obsolete. How could that be anything less than an existential threat to a company that is reliant on license fees?
They're not immune from losing, but they are immune from the technical definition of disruption (from The Innovator's Dilemma, where "sustaining innovation" also come from).
But you're right, pay-per-use is different from license fees (though they do try to approximate volume), and scalability is also different.
Oracle have also been shifting to the cloud, so they are scalable. I'm assuming they also modify their fee structure to make sense with SaaS, i.e. charge-per-usage.
In the following, I'm arguing in good faith not just to score points or win but seeking insight, and I hope you are too: I'm not sure if that transition matters.
It might matter, if Oracle can't adjust to it: organisationally (internally), can't sell in this new way (e.g. sales pitches and sales schedules must change), etc. Whereas, for AWS, it's business as usual, their systems and people all deal with it, have overcome the problems, and have been optimising it, and continue to.
It is a big change, and it's not clear that Oracle will do it smoothly, although they have made big changes before.
But here we return to disruptive innovation vs. sustaining innovation... does this change in scalability and pricing structure make a difference to the business relationship?
For example, if it became a smaller monthly fee rather than larger annual contract, and a different person lower in the customer organization was in charge of it, it could make a huge difference. Different contact, different stakes, different metrics. OTOH, if the total is similar, they'd probably have the same person handle it the same as before.
I'll be honest, it seems disingenuous to argue that Oracle is immune to disruption in a debate about whether Amazon poses an existential threat, and then later reveal that you were using a specific, hidden meaning of the word -- especially when the plain, obvious meaning is directly applicable to the discussion. I will try to assume positive intent.
The last thing I'll say is a bit of advice: In future debates, it would be useful to indicate at the beginning when you're using a specific, non-obvious definition of a word, especially when the broadly accepted meaning is directly consistent with the debate topic, and especially when the definition you are using has little to do with the overall discussion. It will save time and will help avoid misunderstandings that damage your credibility among your counterparts in the discussion.
Sorry, no ill intent, I thought the technical meaning was common knowlwdge on HN, since the technical meaning is where the term originated. I was actually afraid I was being condescending by referencing the source.
But I only introduced it in my 4th comment in the debate, and clarified it in the next one, when I sensed there might be a misunderstanding. I didn't use it at the "beginning" of the debate.
The definition is central to the debate. I see now it's hard for us to discuss this, as disruptive vs. sustaining determines whether the incumbent (Oracle) or entrant (AWS) wins. The debate turns on the distinction.
"Disruption" is just a theory, doesn't have a perfect track record, and I sense you're not interested in it, but the original research on the hard disk drive market was pretty influential, about how some dramatic tech changes in e.g. the magnetic surface made no difference to who led the market ("sustaining"), while changes in form-factor (8", 5.25", 3.5" etc) "disrupted" the market by changing who bought them and why... and also changed who led the market. Wikipedia on "the innovator's dilemma" is a pretty good summary, IIRC.
It's a shame a misunderstanding has prevented you from participating in a discussion. I thought we were reaching the crux of the issue... oh well. It was interesting to me, at least.
After sleeping on it, I agree it is an existential threat.
Disruption usually happens from the low-end, but in this case, Oracle is being threatened from the top-end. That's terrifying. Oracle must address these security concerns.
It's just that I'm not sure the existential threat will succeed (because I'm not sure cloud scalability/pricing will change the reasons to buy Oracle - I find that an interesting question).
BTW When people quote a term, it's to alert you to usage of that term.
Oracle has direct cloud Iaas, PaaS, etc. offerings, and among AWS’s offerings are cloud RDB offerings that compete with Oracle licenses (including derivatives of Oracle MySQL.)
They are absolutely competitors in the spaces of each firm's core business.
Note: I worked at Oracle (although not as a manager), and am still somewhat involved in the ecosystem.
Oracle wants to move most (or even all) of its business to SaaS/Paas/IaaS/WTFaaS. They believe the aaS models are fundamentally more remunerative than on-premises, and can be controlled in an easier way, massively reducing their costs and increasing their lock-in. They also see customers eager to ditch their own datacentres, and believe that they will eventually migrate away to cloud-native services. That means the Amazon juggernaut will eat up more and more of their profits as it expands up the stack (starting from RDS).
So Larry has been steering everything pretty hard, for about 7 or 8 years now, to be sold "aaS". Their offering is pretty large by now (the only thing they don't provide, IIRC, is raw VMs). Entire product lines have been rewritten almost from scratch. They literally won't sell you on-prem licenses unless you also buy "cloud credits". They now compete (or rather try to compete) directly with Amazon in a lot of areas (authentication, PaaS, database, orchestration etc), and are busy "going down the stack" while Bezos is "going up".
Oracle is pulling all the stops to execute on Larry's diktat. They've even condemned the entrenched partner ecosystem, which was one of their strengths, to basically retool into integrators or die, fighting for scraps.
> Oracle also accused Amazon of breaking the rules by hiring two senior DoD staff, Deap Ubhi and Anthony DeMartino, who were involved in the JEDI procurement process. Ubhi is described as "lead PM". A third name is redacted in the publicly released filing.
More important is the next paragraph (apparent typo on the name left in):
> It is alleged that Amazon failed to notify the DoD that it was offering Udhi, the DoD's lead for the procurement project, a job. AWS has always denied any wrongdoing.
If it can be shown that Ubhi influenced the decision before departing for Amazon, then there are some serious legal problems in store for both Ubhi and Amazon.
> If it can be shown that Ubhi influenced the decision before departing for Amazon, then there are some serious legal problems in store for both Ubhi and Amazon.
Honest questions coming out of legal ignorance: isn't this just normal m.o. in the USA these days? Isn't that what "revolving door"[0] is all about? Is this really a legal issue for the participants in this scheme at-large? If not, what did the AWS hires do differently that was more egregious?
Revolving door is a concept but various bodies have different rules limiting when and how it is acceptable.
Oracle is alleging that AWS/DoD violated those rules, now it’s up to the court to decide if they did.
Edit: I’m sure that Wikipedia link doesn’t cover the full universe of rules, but it does say:
> "Under current law, government officials who make contracting decisions must either wait a year before joining a military contractor or, if they want to switch immediately, must start in an affiliate or division unrelated to their government work. One big loophole is that these restrictions do not apply to many high-level policy makers..., who can join corporations or their boards without waiting." [21]
> "Under current law, government officials who make contracting decisions must either wait a year before joining a military contractor or, if they want to switch immediately, must start in an affiliate or division unrelated to their government work. One big loophole is that these restrictions do not apply to many high-level policy makers..., who can join corporations or their boards without waiting."
From your link (didn't bother following it to its source, but it's a correct summary). The President's Cabinet members can freely go back and forth between industry and policymaking positions (well, up to confirmation after nomination which will limit things). But a member of the military/civil service is limited in how quickly they can go to a contractor and in what role they can work.
So my middle class taxes are going to Amazon development.
I don't understand why there is no anti/limited government party in the (modern) United States. The two parties are both advocates for this kind of spending.
There doesn't seem to be any advocates for the masses.
What's funny about this is that Oracle has a long and colorful history of doing the exact same thing[1] that they are accusing AWS of doing in this case. It feels like Oracle is just unhappy that they got outmaneuvered at a game they practically invented.
> Oracle's filing said that US "warfighters and taxpayers have a vested interest in obtaining the best services through lawful, competitive means... Instead, DoD (with AWS's help) has delivered a conflict-ridden mess in which hundreds of contractors expressed an interest in JEDI, over 60 responded to requests for information, yet only the two largest global cloud providers can clear the qualification gates."
Changing the thing being procured shines a light on just how bad this argument is: "Over 60 potential manufacturers declared an interest in designing and building the latest jet fighter, yet all but two were excluded by the qualification gates of being able to provide a product capable of remaining airborne. Hundreds of potential suppliers could provide the odd wing or ejector seat, which put together stand a good chance of resembling a plane."
(Realisations that this is actually maddeningly close to how some jet fighters are built are welcomed)
At a previous job, our company and three other companies operated similar programs, supporting the military. Call it company A, company B, and company C. Our company (company A) were doing a really good job. Our metrics were outstanding, we had the enthusiastic support of 3 consecutive commanders of the unit we were supporting, life was grand. Company B was also doing a great job, and they were better than us at certain things, but as a whole, we were slightly better. Company C suuuucked. Their metrics were shit, so they started gaming the metrics. Their metrics still sucked. The people in the unit we were supporting were well aware of their bullshit, and they regularly got dressed down in front of all us. They were absolute dogshit.
Imagine Firefox vs Chrome vs IE6. In the general case, you want Firefox, (Company A) unless you spend a significant amount of your time using Google Docs, in which case you want Chrome. (Company B) You don't want IE6.
Anyway, the Pentagon needed to select one of these programs to build into a long term DoD program, as opposed to a live in theater prototype, so they commission a study to evaluate the three programs. They hired an outside contractor to perform this study. The company they hired to perform this study was Company C. Company C's study showed that Company C's product was the most effective one. Their numbers were straight up lies- they literally made up bullshit and wrote it down on a piece of paper and submitted it to the Pentagon as a formal document. The unit we were attached to was livid, and wrote up a formal complaint to the Pentagon. This was ignored.
Our program was canceled, I was laid off. I don't work in defense anymore.
US Government agencies tailor bids to favor companies every single day. We call them "wired bids."
Government agencies are made up of people. Those people develop relationships with their providers, when they like them. They've been around long enough to have seen good providers and bad providers. When they find a good one, they hang on to them for a while.
Eventually someone above them shakes things up a little, and a new provider upsets the incumbent and takes over the work. After a few tries - or the first if they're lucky - they'll find another good provider, and the cycle starts over.
That's why bidding on a job isn't where the opportunity is won. You almost always win or lose the job before the solicitation is released. You win by building relationships with the people that make up these agencies. Do research, set up meetings, go talk to people, figure out what gives them heartburn, and develop a solution to their problems.
That solution becomes the framework for the new solicitation. You have then influenced the requirements. And congratulations, the bid has now been tailored to favor your company. You will probably win!
Not if the alternative is Amazon. There's old-school corporate bad, and there's the terrifyingly horrendous. Never mind that Amazon is a company that's destroying most of retail without direct cache infusions from the US government (other than tax relief), it's a company that terribly mistreats its employees and eavesdrops on its customers.
Great metaphor. To play devil's advocate: if only one or two companies can satisfactorily make cars, should other less capable/affluent manufactures accept defeat and hand over the automobile market to them? Look for another line of business to break into?
There's no question that Amazon and Google are the most capable (perhaps the only viable) choices for a contract of this scope. But giving them 10 billion dollars and relative corporate invinciblity (Amazon will not go under unless the US government decides so/collapses) is just exacerbating the problem. A government concerned with preserving the free market (and thus the prosperity/happiness of their citizens) should be stimulating smaller competitors, not funneling all available resources to a small handful of megacorps they're in bed with.
The trouble is, the government gets constant complaints about services not working, or not being up to par with the public. The government should probably not be a proving ground for manufacturers that have not successfully delivered before. Instead, private enterprise can do that, and success with private enterprises should be a qualifying event for future public delivery.
An interesting observation, though, might be that consumer/public safety are opposed in some form to a truly competitive market. many companies could build vehicles that couldn't meet emissions or crash test requirements. But, if we want to guarantee those things, then it necessarily reduces the competitive pool to those enterprises willing to invest to be able to meet those standards. It is probably not of primary significance, but I like how policies aimed at saving consumers/the public from harm can also cause the public harm in the form of reduced competition and increased income inequality down the road.
> consumer/public safety are opposed in some form to a truly competitive market.
I had a relative in the chemical industry, and she was always bitching about this: "small players can never match the big ones in certain areas, because they can not afford the red-tape to certify this or that". On the other hand, these regulations exist for a reason - most legislators don't wake up in the morning excited to ban some obscure compound. I'd rather have fewer jobs and money, if I can be sure I won't be poisoned so that someone else can make a quick buck.
> There's no question that Amazon and Google are the most capable (perhaps the only viable) choices for a contract of this scope. But giving them 10 billion dollars and relative corporate invinciblity (Amazon will not go under unless the US government decides so/collapses) is just exacerbating the problem.
it may be exacerbating a problem within the FTC’s domain of responsibility for market competition, but defense (or government more generally) procurement, even in an ideal world , would not involve systematic accommodation for providers less capable of delivering the requirements.
> A government concerned with preserving the free market (and thus the prosperity/happiness of their citizens) should be stimulating smaller competitors
A government concerned with maintaining free markets will, by definition, buy things from the market based on most cost effective fit for needs, not choosing which vendors should be promoted on other grounds.
What you are suggesting may be promoting some concept of healthy markets at the direct expense of free markets by direct government intervention, but it's not helping free markets.
In all fairness Oracle does have a superior cloud infrastructure, their cloud control code and user application code does not run on the same physical hardware.
51 comments
[ 5.3 ms ] story [ 135 ms ] threadAnyone whose used AWS/Azure/Oracle Cloud will tell you why that idea is problematic. Particularly when dealing with a hardened government version of these cloud environments.
The only potential corruption their complaint indicates is the hiring of a decision maker at the DoD. It is a bad look. I'd need to see the timeline/understand more about it to form an opinion.
Having more then 1 company that provides the service could help preventing vendor locking and provide better proces c in future.
Edit: To expand a bit further: https://aws.amazon.com/getting-started/projects/migrate-orac...
AWS is a cloud platform provider.
Unless one (or both?) have changed while I wasn't looking... but even then, is the Cloud now so central to Oracle that its existence depends on it?
EDIT thanks for the link. "Oracle data warehouse" isn't Oracle's central business. Also, I'm not a fan of Oracle, but I doubt it's so easy to steal their customers. Finally, how can the loss of one govt contract be so harmful? Few of their customers, not even govt ones, are so security conscious... BTW I realize you're just reporting their claim, not making it.
As an aside, I regret not disclosing earlier that I'm an engineer at Amazon. I did not expect to participate in this discussion this much, and also, it just didn't occur to me. My apologies.
From your first link, compatibility is hard: AWS SCT clearly marks any code that it cannot convert so that you can manually convert it.
I would expect O to be upset that they don't even get a look in to an incredibly lucrative contract, and argue it's bad for the govt's well-being to exclude them. I can't see how or why they'd argue it threatens their existence...
Oh, so you are making the claim! Your comment makes a lot more sense now.
BTW congrats on winning the contract, that's amazing!
The thesis of my argument is that, in my opinion, Amazon represents a true threat to Oracle, and that moves like this filing should be examined through that lens.
For the record, my opinions are my own and I have no knowledge of this situation or AWS decision making that is not public.
direct competitors != existential threat
But you're describing a "disruption": AWS's offering is comparable yet far cheaper. Which sounds like an "existential threat".
Oracle should be immune to this disruption: the same people are doing the same thing to the same data for the same purpose - so, the cloud is a "sustaining" innovation in this context.
Plus, Oracle is a sales company, not an engineering company, and I don't think AWS is oriented to compete that way.
You've softened your thesis to a "threat", and yes I agree of course it is. And yes, their competitive situation will inform their actions.
But... $10 billion dollars is reason enough to file in itself, right? They have huge departments devoted to just this kind of thing. And recall the fuss they made vs Google over Java, for much less (though TBF, that had strategic advantages for them).
I still don't see how this particular contract would have knock-on consequences for Oracle, making it an existential threat. Maybe I'm missing something.
> Oracle should be immune to this disruption: the same people are doing the same thing to the same data for the same purpose - so, the cloud is a "sustaining" innovation in this context.
The idea that any company is immune to disruption seems obviously wrong to me. But in this case, it seems very clear. You're correct that people use databases the same way for the same purpose on cloud vs self host -- no one disputes that. What the cloud dramatically alters, among many things, is pay-what-you-use pricing and scalability. AWS database offerings render license fees obsolete. How could that be anything less than an existential threat to a company that is reliant on license fees?
But you're right, pay-per-use is different from license fees (though they do try to approximate volume), and scalability is also different.
Oracle have also been shifting to the cloud, so they are scalable. I'm assuming they also modify their fee structure to make sense with SaaS, i.e. charge-per-usage.
In the following, I'm arguing in good faith not just to score points or win but seeking insight, and I hope you are too: I'm not sure if that transition matters.
It might matter, if Oracle can't adjust to it: organisationally (internally), can't sell in this new way (e.g. sales pitches and sales schedules must change), etc. Whereas, for AWS, it's business as usual, their systems and people all deal with it, have overcome the problems, and have been optimising it, and continue to.
It is a big change, and it's not clear that Oracle will do it smoothly, although they have made big changes before.
But here we return to disruptive innovation vs. sustaining innovation... does this change in scalability and pricing structure make a difference to the business relationship?
For example, if it became a smaller monthly fee rather than larger annual contract, and a different person lower in the customer organization was in charge of it, it could make a huge difference. Different contact, different stakes, different metrics. OTOH, if the total is similar, they'd probably have the same person handle it the same as before.
The last thing I'll say is a bit of advice: In future debates, it would be useful to indicate at the beginning when you're using a specific, non-obvious definition of a word, especially when the broadly accepted meaning is directly consistent with the debate topic, and especially when the definition you are using has little to do with the overall discussion. It will save time and will help avoid misunderstandings that damage your credibility among your counterparts in the discussion.
But I only introduced it in my 4th comment in the debate, and clarified it in the next one, when I sensed there might be a misunderstanding. I didn't use it at the "beginning" of the debate.
The definition is central to the debate. I see now it's hard for us to discuss this, as disruptive vs. sustaining determines whether the incumbent (Oracle) or entrant (AWS) wins. The debate turns on the distinction.
"Disruption" is just a theory, doesn't have a perfect track record, and I sense you're not interested in it, but the original research on the hard disk drive market was pretty influential, about how some dramatic tech changes in e.g. the magnetic surface made no difference to who led the market ("sustaining"), while changes in form-factor (8", 5.25", 3.5" etc) "disrupted" the market by changing who bought them and why... and also changed who led the market. Wikipedia on "the innovator's dilemma" is a pretty good summary, IIRC.
It's a shame a misunderstanding has prevented you from participating in a discussion. I thought we were reaching the crux of the issue... oh well. It was interesting to me, at least.
Disruption usually happens from the low-end, but in this case, Oracle is being threatened from the top-end. That's terrifying. Oracle must address these security concerns.
It's just that I'm not sure the existential threat will succeed (because I'm not sure cloud scalability/pricing will change the reasons to buy Oracle - I find that an interesting question).
BTW When people quote a term, it's to alert you to usage of that term.
They are absolutely competitors in the spaces of each firm's core business.
Oracle wants to move most (or even all) of its business to SaaS/Paas/IaaS/WTFaaS. They believe the aaS models are fundamentally more remunerative than on-premises, and can be controlled in an easier way, massively reducing their costs and increasing their lock-in. They also see customers eager to ditch their own datacentres, and believe that they will eventually migrate away to cloud-native services. That means the Amazon juggernaut will eat up more and more of their profits as it expands up the stack (starting from RDS).
So Larry has been steering everything pretty hard, for about 7 or 8 years now, to be sold "aaS". Their offering is pretty large by now (the only thing they don't provide, IIRC, is raw VMs). Entire product lines have been rewritten almost from scratch. They literally won't sell you on-prem licenses unless you also buy "cloud credits". They now compete (or rather try to compete) directly with Amazon in a lot of areas (authentication, PaaS, database, orchestration etc), and are busy "going down the stack" while Bezos is "going up".
Oracle is pulling all the stops to execute on Larry's diktat. They've even condemned the entrenched partner ecosystem, which was one of their strengths, to basically retool into integrators or die, fighting for scraps.
> It is alleged that Amazon failed to notify the DoD that it was offering Udhi, the DoD's lead for the procurement project, a job. AWS has always denied any wrongdoing.
If it can be shown that Ubhi influenced the decision before departing for Amazon, then there are some serious legal problems in store for both Ubhi and Amazon.
Honest questions coming out of legal ignorance: isn't this just normal m.o. in the USA these days? Isn't that what "revolving door"[0] is all about? Is this really a legal issue for the participants in this scheme at-large? If not, what did the AWS hires do differently that was more egregious?
[0] https://en.wikipedia.org/wiki/Revolving_door_(politics)
Oracle is alleging that AWS/DoD violated those rules, now it’s up to the court to decide if they did.
Edit: I’m sure that Wikipedia link doesn’t cover the full universe of rules, but it does say:
> "Under current law, government officials who make contracting decisions must either wait a year before joining a military contractor or, if they want to switch immediately, must start in an affiliate or division unrelated to their government work. One big loophole is that these restrictions do not apply to many high-level policy makers..., who can join corporations or their boards without waiting." [21]
From your link (didn't bother following it to its source, but it's a correct summary). The President's Cabinet members can freely go back and forth between industry and policymaking positions (well, up to confirmation after nomination which will limit things). But a member of the military/civil service is limited in how quickly they can go to a contractor and in what role they can work.
I don't understand why there is no anti/limited government party in the (modern) United States. The two parties are both advocates for this kind of spending.
There doesn't seem to be any advocates for the masses.
[1]: https://www.latimes.com/archives/la-xpm-2002-may-04-me-oracl...
Changing the thing being procured shines a light on just how bad this argument is: "Over 60 potential manufacturers declared an interest in designing and building the latest jet fighter, yet all but two were excluded by the qualification gates of being able to provide a product capable of remaining airborne. Hundreds of potential suppliers could provide the odd wing or ejector seat, which put together stand a good chance of resembling a plane."
(Realisations that this is actually maddeningly close to how some jet fighters are built are welcomed)
A government agency tailoring a bid to favor or not favor any company is entirely illegitimate (and illegal).
At a previous job, our company and three other companies operated similar programs, supporting the military. Call it company A, company B, and company C. Our company (company A) were doing a really good job. Our metrics were outstanding, we had the enthusiastic support of 3 consecutive commanders of the unit we were supporting, life was grand. Company B was also doing a great job, and they were better than us at certain things, but as a whole, we were slightly better. Company C suuuucked. Their metrics were shit, so they started gaming the metrics. Their metrics still sucked. The people in the unit we were supporting were well aware of their bullshit, and they regularly got dressed down in front of all us. They were absolute dogshit.
Imagine Firefox vs Chrome vs IE6. In the general case, you want Firefox, (Company A) unless you spend a significant amount of your time using Google Docs, in which case you want Chrome. (Company B) You don't want IE6.
Anyway, the Pentagon needed to select one of these programs to build into a long term DoD program, as opposed to a live in theater prototype, so they commission a study to evaluate the three programs. They hired an outside contractor to perform this study. The company they hired to perform this study was Company C. Company C's study showed that Company C's product was the most effective one. Their numbers were straight up lies- they literally made up bullshit and wrote it down on a piece of paper and submitted it to the Pentagon as a formal document. The unit we were attached to was livid, and wrote up a formal complaint to the Pentagon. This was ignored.
Our program was canceled, I was laid off. I don't work in defense anymore.
Government agencies are made up of people. Those people develop relationships with their providers, when they like them. They've been around long enough to have seen good providers and bad providers. When they find a good one, they hang on to them for a while.
Eventually someone above them shakes things up a little, and a new provider upsets the incumbent and takes over the work. After a few tries - or the first if they're lucky - they'll find another good provider, and the cycle starts over.
That's why bidding on a job isn't where the opportunity is won. You almost always win or lose the job before the solicitation is released. You win by building relationships with the people that make up these agencies. Do research, set up meetings, go talk to people, figure out what gives them heartburn, and develop a solution to their problems.
That solution becomes the framework for the new solicitation. You have then influenced the requirements. And congratulations, the bid has now been tailored to favor your company. You will probably win!
There's no question that Amazon and Google are the most capable (perhaps the only viable) choices for a contract of this scope. But giving them 10 billion dollars and relative corporate invinciblity (Amazon will not go under unless the US government decides so/collapses) is just exacerbating the problem. A government concerned with preserving the free market (and thus the prosperity/happiness of their citizens) should be stimulating smaller competitors, not funneling all available resources to a small handful of megacorps they're in bed with.
An interesting observation, though, might be that consumer/public safety are opposed in some form to a truly competitive market. many companies could build vehicles that couldn't meet emissions or crash test requirements. But, if we want to guarantee those things, then it necessarily reduces the competitive pool to those enterprises willing to invest to be able to meet those standards. It is probably not of primary significance, but I like how policies aimed at saving consumers/the public from harm can also cause the public harm in the form of reduced competition and increased income inequality down the road.
I had a relative in the chemical industry, and she was always bitching about this: "small players can never match the big ones in certain areas, because they can not afford the red-tape to certify this or that". On the other hand, these regulations exist for a reason - most legislators don't wake up in the morning excited to ban some obscure compound. I'd rather have fewer jobs and money, if I can be sure I won't be poisoned so that someone else can make a quick buck.
it may be exacerbating a problem within the FTC’s domain of responsibility for market competition, but defense (or government more generally) procurement, even in an ideal world , would not involve systematic accommodation for providers less capable of delivering the requirements.
> A government concerned with preserving the free market (and thus the prosperity/happiness of their citizens) should be stimulating smaller competitors
A government concerned with maintaining free markets will, by definition, buy things from the market based on most cost effective fit for needs, not choosing which vendors should be promoted on other grounds.
What you are suggesting may be promoting some concept of healthy markets at the direct expense of free markets by direct government intervention, but it's not helping free markets.