I see too many companies using AWS for stuff they could fairly easily self-host.
Obviously, web-hosting stuff is a little different. AWS is a lot more than just web hosting: it's a cloud computing provider.
As you do with AWS...
> and created our env
As you do with AWS...
> And it was super easy to move from one provider to another.
Which is easier on cloud hosting than it ever was when dealing with colocation providers.
AWS provides a number of SAAS offerings which, if you make use of them, tie you to Amazon specifically, as opposed to an IAAS offering which is likely to be more easily portable to another provider.
Having to get approval for a capital expense was a filter that prevented teams with zero ops experience from trying to just brute-force google their way through.
The teams I see the most often, in pre-cloud days got nothing done because touching a server meant going through experienced ops teams that could conduct architecture reviews, security reviews and maintenance planning but which didn't have any motivation or incentives to actually launch new products. Cloud means these teams can get things done, at the expense of not having any thought put to architecture, security, or maintenance.
Sure, it's not hard, but it's not possible for them.