But I got your point on politics, that is something pretty unpredictable.
But, it is a maximum of $10Billion ...over 10 years. Th at is the key thing to focus on. I'm sure both companies have private sector customers right now that pay a quarter to half-billion a year right now.
I think the greater damage is optics vs financial..and optics can cost a lot.
Let's say AWS makes $30-Billion this year. So, losing $1-billion this year is 3.33%? (Though I'm sure the contract starts not this year?)...How much will AWS be making in 10 years?
Once all of the pentagon is integrated with Azure, after the 10 years are up, it will be very unlikely for Pentagon to want to switch, not only due to costs, but everyone will have been trained specifically for handling Azure.
This gives Microsoft a massive lead in terms of government infrastructure and all but guarantees we'll be seeing more and more contracts handed to Azure.
https://www.wired.com/2010/08/hp-holds-navy-network-hostage/
Politics nowadays are just way too unpredictable and destructive. Just best not to be anywhere near it when, say, cost overruns on this project happen and you still haven't delivered anything the DoD is actually using.
How? AWS already has 3 entire regions dedicated to the US government and those aren't going away anytime soon.
Note: Pentagon != all US govt agencies
it is a maximum of $10Billion ...over 10 years
... because government contracts never go over original estimates?But, government contracts are pretty beholden to a tight, static budget. Once the money is gone, it's gone. You can't necessarily go ask for additional taxes and budget for the year. Many times, you literally cannot go over budget no matter what. The pay-as-you go, variable pricing you get in the cloud, is actually a challenge in government budget.
In the private sector, when you go over budget..funds can be found elsewhere in the company, etc.
Ask Frank Underwood says: When the tits are that big, everybody gets in line.
And 10 billion over 10 years is what? Maybe 3 to 5% of Amazon's yearly revenue? You really want to be at the epicenter of the shitstorm that will happen if the project goes over budget and you still haven't delivered anything anyone in the DoD actually uses? (Maybe I shouldn't have said "if the project goes over" but "when"?) Especially being a big tech firm nowadays?
And you're gonna step into all that unnecessary drama for access to a 10 year / 10 billion project?
I don't think so. Not for a billion a year in revenue when you're already sitting at over 230 billion a year in revenue. (Or even just the over 30 billion a year that AWS alone is sitting at.)
Admit it man, politics being what they are these days, particularly for big tech firms, it's just not worth it.
That's the point.
It's the shitstorms that generate the tweetstorms that impact their larger bottom line.
(They seemed to like the idea of New York a little bit. Probably because of multimodal transport connections to NoVA and the Europe/Africa side of the world. But even that fell through. And as far as I can understand, that too was because of politics.)
Also, Amazon has a tremendous footprint here already. Their largest region is located in Ashburn (~25 miles east), CIA is roughly 5 miles for Crystal City, Herndon and Ballston have large AWS sales/engineering offices, and Richmond (roughly 1.5 hours south) hosts one of their largest distribution centers on the east coast. In short, they were committed to this area long before HQ2, let alone JEDI. They cannot be happy about losing the contract award, but it was only one of many things Amazon/AWS are going in NoVA.
They just seemed to let go of New York and JEDI wayyy too easily for being one of the wealthiest companies in the world. When someone like that bows out of such large projects so easily, something's off. And I have to take it at face value that they maybe just want revenue and not politics. Because no other reasons seem to present themselves here.