In Praise of Negative Reviews

YestedayRich Adin complained specifically about customer reviews on a book which features a positive quote from him in the Book Description, and more generally about negative reviews written by the uneducated. In it, he asks a question and raises a problem. Here is the question: “If you didn’t read the book, why rate it? And why give it a 1-star rating?” I love negative reviews. The favorable reviews could be written by sock puppets of the author or maker, but negative reviews are written by people who feel strongly about something and I’d like to know why. If I think they are idiots, those negative reviews add to my confidence that I am making a good choice. But if a product has a huge number of mostly positive reviews, a warning might be embedded in those negative reviews that would be hard to extract from the positives.

In the general case, a person who attempted to read a book and failed might rate a book to warn others about their experience. You might do this with a restaurant you didn’t eat at, if you made a reservation, arrived slightly early, and were still not seated (or had been seated but not yet placed an order, or whatever) an hour after the time of the reservation. There could be any number of reasons why you were unsuccessful at eating (failed to adhere to a dress code, restaurant closed by the health department, visible rat dung etc.) and a person contemplating eating at that restaurant might want to know which of those reasons applied -- but the failure to eat a meal there is no reason why you shouldn’t post a review of your extremely negative experience.

If I were selling an historical novel about women and their intimate relationships, I’d be looking at the Romance Novel Consuming Audience and slavering. Copiously. But I’d also know better than to sell a book without an emotionally satisfying ending to this audience. There are romance novel series, but each entry has an emotionally satisfying ending -- if not an Ever After, a Consummation that was Devoutly Wished. The reviews -- positive and negative -- of the book Rich Adin uses as his example are generally quite clear that an HEA is not something that the book will provide, yet it attracted a Romance Novel Consuming Audience that expected on. The product was not described correctly. Usually, books like this come with warning labels (“tearjerker” and “anguish” and “stark” being just a few examples of words that could have provided adequate warning). If the ride doesn’t say there are strobe lights and someone has a seizure, they’re going to be pissed and tell all of their neuro-diverse friends.

Here are a few examples of the negatives reviews of _Sentence to Marriage_.  A two star review by reading mom says “I wish someone would have mentioned how the sadness just keeps going on.” She notes that it is well-written, the characters are well developed and the story probably accurate to history but objects to the overall length, and “thought the details of minor events went on too long with not contributing to the story line”. A one star review by chris83 says, “Sentence of marriage should have been titled sentence of agony. This book was not a romance.” And there are many, many more in which the fact it is NOT a romance and does not have an HEA and is incredibly depressing are the primary features of the review. These reviewers are emphatic that they wished they’d given up sooner and gone on to read something else -- exactly the kind of information this kind of reader needs and would want them to have, and which should have been provided in the Book Description.

I want to thank Rich Adin for correctly spelling the idiom in this sentence: “As I said before, the problem is giving free rein to anonymous reviewers who are unknowledgeable about the book being reviewed.” I disagree, however, with his assessment of the problem. While it would be nice if all reviews were connected to identities that we could track (might help with the sock puppet problem) and thus attach reputations to, Adin has significantly misunderstood the tradeoffs involved in providing product reviews on e-commerce websites. has been providing reader reviews since the first few months that it was open (I don’t actually know if they were part of the initial web site rollout). The reviews were a very simple feature to implement from a coding perspective and inexpensive in terms of computing resources required to include in a detail page. The decision to have little or no security, checking, review, etc. of reviews was made at a high level at the beginning and only modified over time in the face of overwhelming need because reviews: are cheap for the site to provide only as long as there is no (later, little) QA on the reviews; make the site attractive to non-customers who then become customers; make the site sticky to customers who then buy more; provide a barrier to entry to competitors whose volume of reviews is necessarily smaller.

Adin’s proposal (no anonymous reviews) would violate several of these. Mandatory identity checking in particular would reduce the volume of reviews for a host of reasons, one of which is particularly salient in light of the recent problem some reviewers at goodreads have been experiencing: harassment of people who provide negative reviews. In any event, has no motivation to take on the problem Adin has with the review, particularly as long as so many customers are finding the negative reviews so helpful. wants people to buy the kind of books they’ll want to buy more of -- not the kind of books that will make them want to take a break from reading for a long while because it was unexpectedly harrowing. Adin may want people to be challenged by their reading. I don’t think cares.

The person who is using reviews for their ostensible purpose -- separating the wheat from the chaff, as Rich Adin biblically put it -- does have a responsibility to use those reviews appropriately. The comments thread in response to Adin’s post make it quite clear that most people have known all along that Reviews Need Translation. As fjtorres put it, “Part of being an educated consumer is knowing how to parse and deconstruct reviews.” Using the universal experience of movie review translation, fjtorres adds: “I no more accept an annointed expert’s opinion as gospel truth than I accepted it from a clueless nobody, yet both serve a purpose.”

image by cnewtoncom

9 Comments on In Praise of Negative Reviews

  1. Yes! I made the comment on the Rich Adin post that what’s most important in a review is specificity– say exactly what you liked and what you didn’t like! You examples with disappointed romance readers are a perfect illustration. Someone who’s looking for literary fiction and not romance would know from that description that the book is a good bet for them. ANNE KARENINA is never going to sell in the romance department, in spite of the love story in it because it’s the ultimate sad ending.

  2. Hear! Hear! I often look at the 1-star reviews to see if they’re complaining about something I don’t mind or actually prefer.

  3. Name (required) // 8 August, 2012 at 3:01 am // Reply

    I do not look at the reviews often. But when I do, in order to decide whether I want to read the book, I often find negative reviews much more informative than positive ones. Yes, they have the unfortunate side-effect of lowering the overall score, but they tell you important things. I the past I *have* decided that I was going to like the book based on negative reviews.

  4. Why this aut-aut position? Can’t we have both, anonymous AND accountable reviews in the same site?
    And let’s make three: one section for customer anonymous unmoderated reviews, one for identified educated readers and one for professional reviewers…

    • Doesn’t Amazon do it that way? I’ve assumed that the “Real Name” designation floating near certain usernames is intended so that it’s easy for people that want the opinion of a well-known individual to know it’s not an impostor.

  5. I think you wrote a fine rebuttal but I would like to make one clarification of my view: I am not opposed to negative reviews. What I am opposed to are negative (and positive) reviews that are anonymous and that are based on fantasy. It is fine to complain that the book description misled the reviewer into thinking it was a romance novel when it is not, but it is not OK to to rate the book 1 star or 5 stars if you haven’t read it or attempted to read it. Similarly, I think it is wrong to rate a book 1 star solely because you do not like the price being charged, especially when you haven’t bought the book, yet disguise your real reason for the rating by falsely claiming, for example, that the book is poorly edited when in reality it is very well edited. Anonymity encourages too many people to simply post whatever comes to mind “as a lark.” In the case of the book I cited, at one bookseller the book had close to 500 “reviews,” most of whicvh are anonymous. That is a large number of reviews to have to wade through and interpret.

    • I just have two words in response to the request for real names. Harriet Klausner. There’s a helpful real shame reviewer. Not.

    • That’s not really the reviewer’s fault, though – it’s a limitation of Amazon’s website. There’s no choice about giving a star rating, and in cases like this, it may well be unfair. There’s the flip side of this as well – I’ve read quite a number of reviews where no star or even a minus rating would have been a more accurate reflection of my reaction to it.

      However, in this case, my sympathies are with the reader more than the author. If the book description was misleading, I can’t see how the author can expect this not to be a factor; they have to take some responsibility for their book description being at least part of the reason that the reader was misled and thus disappointed. Many blurbs aren’t even written by the author – they’re written by some publisher’s lackey – so if that makes the author more proactive about writing their own descriptions, it can only be a good thing. I do, however, agree with you that people who give low ratings just because of the price or because it wasn’t delivered on time or another reason that really is outside the author’s control are just annoying, and I do feel that there should be a mechanism in place to have their impact on the overall rating removed. I don’t agree that the majority of reviewers post bad reviews as a ‘lark’ or make up reasons for giving a poor review – reviews take time to write, and few trolls would have the patience to keep that up for long. In general, I find that the reviews (particularly in large numbers) are a pretty good indication of the book’s quality.

2 Trackbacks & Pingbacks

  1. Stumbling Over Chaos :: Lack of lemurs lowers linkity
  2. The positive side of negativity | Readers' Salon

Leave a comment

Your email address will not be published.