Google is good at launching products, they're not good at maturing them. Chrome is a fantastic piece of software, and it fulfills it's purpose for google quite well, but it's not a product in the way that AWS or the kindle is a product.
Google is good at development, excellent in fact. But they get stuck in the hard middle parts of product development. They fail to monetize products well and they fail to mature products into their full potential. Partly that's because product development at google is basically just play-time. As I said, revenue for google derives from a tiny subset of products (search and network almost exclusively), everything else is just a side project. As such, when a product doesn't meet its potential or doesn't bring in billions in revenue that is not an issue at all at google, it's the expected case. In some ways that's good, but in other ways it's bad. Docs/drive has stagnated relative to what it could have been for years.
Amazon is almost diametrically opposite. They productize well, extremely well, but they have a horrid working environment in regard to devs.
Amazon has a strong history of expanding their business and building new highly valuable products with staying power. Google has only expanded a little in terms of actual functional products, most of their additions have been via acquisitions, and they have not done well at monetizing or even supporting many of their new product offerings. Google is primarily search + backbone ISP, that's where almost all of their revenue comes from, everything else falls into a tiny "other" category. Indicative of their utter inability to effectively spinoff new products and high revenue businesses.
The only downside is that the work environment is more than a little dysfunctional.