If by: my "audience", you mean: people in general, you'd be correct.
Sales being my primary focus, and given people often say one thing while doing another, I think the best indicator of whether my book's effective or not is performance on the balance sheet. If many random reviewer/consumer's actions contradict their words - as is often the case - then those reviews are useless to me from a book-selling standpoint.
In the interest of accessing the best information I can, which thereby helps improve sales, I need to remove the randomness, inconsistencies, and hypocrisy humans exemplify to the greatest extent possible. Paying any mind to the words of random, chaotic reviewers on Amazon or Goodreads runs counter to that objective.
Corvid, I'd just like to chime in to say how much I agree with you on all of this. It is not about having a negative opinion about one's audience; rather, it is being realistic about who consumers are as a group. Reading reviews of any product can drive a person mad with the inconsistencies, misinformation, and randomness. There are better ways to receive feedback. If I changed how I wrote to abide by each review as it came I would no longer write what I love and I would have incoherent messes of books that likely would stop selling.
To the OP, I stopped reading all reviews in 2018 because it is a net negative. While I think it is awesome a lot of authors have, I never learned anything useful or actionable from a review. Instead, the randomness, inconsistencies, and hypocrisy Corvid mentioned drove me mad when trying to apply feedback. What one person loves another hates. One person claims one thing, while another claims the opposite (this character is too strong/weak, or I love/hate this). With some reviews, you can't even recognize your own book in what they're saying. Negative reviewers move through a whole series (once again abiding by the behavior Corvid mentioned). Overwhelming positive reviews can do more harm than good if they sound fake or don't go into detail. Some reviews--negative and positive, but particularly positive--are incoherent, so I can't glean anything out of them.
The best way I have found to hear useful feedback has been asking my audience directly. Sending out short surveys in my newsletters is a good method, and I often get great, enthusiastic, detailed responses. I also love getting fanmail from super-fans because they tend to go into detail about the things that really spoke to them, and I get a better insight into how my work is translating to others. Of course, even these methods highlight the inconsistencies. I did a survey once where most of my fanbase said they wouldn't buy a standalone, yet the next book I released (a standalone character prequel to the already released series) became my new bestseller and had high pre-orders. So you have to take even this kind of intimate feedback with a grain of salt.
