"As it turns out, the ones who do it well are rare and far less visible because - like good programmers - their work is a lot harder to notice."
This sums my experience perfectly. It takes a long time to be able to understand who is good at something and who is not.
Pardon my language but many marketers are charlatans. When I first started in marketing I thought the charlatans were the ones who were good (incompetent people [me at that time] are the least competent at recognizing those who are good). Now, after a decade at it I can firmly agree with the statement above and I think it bodes very well for the author.
I believe the 2nd point is the most important to understand when trying to get good at marketing. The rest is ancillary to doing a great job, but having the data is essential to knowing what to do.
I believe great marketers are like methodical scientists. They can put aside their own biases and test their hypothesis and analyze the data. Beyond that, they can look at the data in front of them, identify what they need to know more about, and discover things they did not expect. Through this they can affect change that improves the entire organization.