The problem with editorial analytics

Analytics- a great way of identifying audiences or encouraging the pursuit of clickbait?

In the last 5 years analytics have become a vital part of the career of journalists particularly in the newsroom where they dictate a large part of editorial analytics. In the past when producing a piece of work a journalist would not have anywhere near as much information on their audiences but now they are able to identify where, how long and even how far down the page a user scrolls.
From the use of analytics the writers dictate what kinds of work they produce so the audience is in full control of what they are reading but the information presented in analytics is not always as useful as we may think for example if a reader is scrolling through an article but then clicks off it to go to cook some food it appears to the analytics that the reader has become bored of the article but that is not always the case, so this information is useful but is it always solely what journalists should go by when deciding what to produce?

Overall the key issues that have arisen from the use of analytics are; that creativity is taken from journalists because they are encouraged to chase clickbait as opposed to producing fulfilling work that lives up to the editorial standards of a publication, journalists do not always know how to use the analytics to help their work and the analytics are not always the answer to identifying audiences.