Well a few points below, but probably the most important one is you are right, a pretty graph is useless without context and a person to put it into perspective, which is why doing the analysis and then selling the perspective of how it impacts the business is critical and can help advance a career.
But for example, Wal-Mart. They use the data in near real-time to make a huge amount of decisions, among them is pricing, what is selling, where it is selling best and with a little work they know the price sensitivity for a product in each market allowing them to maximize revenue. I think there are some Wal-Mart devs on here that know far better than I do, if they can comment.
Or another, Ford. I was on a team that built analytics for them that drove their supply chain decisions in near real time based upon a huge number of inputs and it saved them massive amounts of money by reducing supply chain inventory.
Insurance analysis is done via collected and publicly sourced data to determine risk profiles, rates and a host of things. So yes, the data is driving what policies to sell, when its best to sell it and how much they should charge to maximize revenue. In addition market analysis tells them which policies to offer where.
A number of the big box hardware type stores use analytics to determine where to move product and and what time of year to move it to maximize product revenue.
This can go on and on. Silicon Valley is isn't really where I would look for valid references to how this works in the real world, because I think to your point their view is sometimes skewed towards producing the analysis but not converting that into action. But enterprises use data everyday, and are getting better at it to make these decisions. So when you can show a company how to use data they already collect to make a better decision or maximize revenue you can absolutely make an impact and help your own career.