Fascinating rant from Google developer, Steve Yegge who was pervious worked at Amazon. His basic point is that Google doesn't develop with a platform in mind. According to him, every product is developed in a silo with little or no accessibility built in for external developers. This is an amazing insight given Google's constant use of the phrase "open" in its communications (especially within the Android vs. Apple scenario). However, there is one point in the article in which he explains that there is no funding for platform projects and there in lies the rub. Getting funding for internal projects is nearly impossible unless you have a real visionary (aka Steve Jobs, Jeff Bezos) in your company. In a former life, I was a product manager on a search marketing application which used Google, Microsoft and Yahoo web services for reporting and campaign management. By far Google was the best out of the bunch, but there was an arrogance in the way they rolled out changes. Little warning and short time frames were the standard operating procedures. I guess they thought I had thousands of developers at my disposal. To add insult to injury, they begin actually charging for their services. My immediate reaction was, "You are going to charge me for this crap?", but after a year the quality did improve. Roll outs began to be scheduled on a regular basis, documentation improved and service outages went down.
My hypothesis as to why the serviced improved is the fact that they didn't look for internal funding. Users were already finding value in the service so why not charge for it. If there is real value, people will pay for it. The challenge is not finding funding (internal or angel) because you think to have a great idea. Great ideas are worth nothing. The challenge lies in determining at what point is the value is enough to charge customers. It's a lot easier to make the case for resources when there's money flowing into the bank.