2 September 2015 - 6:29 pm

Amazon vs Authors: Review policing

Part of the Amazon is not your friend series.

Any author, whether traditionally or independently published, will tell you that reviews are important when it comes to exposure and sales. They directly influence our bottom line. It’s to be expected, then, that any threat to an author’s ability to get reviews – particularly positive ones  – will cause a stir.

When word started filtering through the internet about Amazon removing legitimate reviews from books, it’s not surprising that authors got upset and vocal about it.

Let’s take a walk through Amazon’s recent review saga.

How it started

Amazon actively encourages reviews from its customers, so that other customers can know how good (or not-good) a particular product is. In 2012, a debacle about unethical reviews and sock puppets occurred: authors were caught paying for reviews, and review providers were creating multiple accounts to leave multiple reviews for the same book. Some authors cut out the middleman and created the sock puppet accounts themselves to spoof good and bad reviews on books, depending on whether the author was a competitor.

As Writer Beware points out, late in 2012, Amazon changed its review policy (and started to enforce it more stringently) to try to stamp out unethical and paid-for reviews. The review criteria are pretty clear (if you know where to look, which is not clear) about reviews being disallowed from anyone who has a financial interest in the product’s success. It also excludes those with a familial or ‘close personal relationship’ with the author.  Amazon appeared to interpret these guidelines pretty broadly when winding up the ban-hammer and aiming it at reviews.

This led to a slew of criticism and complaint, because fans and fellow authors were finding that their reviews were being removed or rejected. Inquiries received a response that stated the review was a violation of the terms and conditions. Pushing the issue (which could mean just asking for more information) resulted in a threat to remove the book from Amazon entirely.

On top of this, there were obvious fakes who were not affected by the wildly swinging ban-hammer (that’s my name for it, and I’m sticking with it).

The situation today

After the end of 2012, the furor died down and it has been pretty quiet since then. Recently, the issue has reared its head again, as more fans and authors trip over this problem. It’s hard to tell if this is a resurgence of the same issue, or if Amazon are enjoying a fresh spree with the ban-hammer.

Some criticism has emerged about the original complaint that brought this issue back into the limelight. However, it’s not an isolated incident. I think there’s enough history and additional instances of the problem that it’s worth taking a look.

It seems that Amazon’s interpretation of ‘close personal relationship’ can mean anyone who has interacted online (it’s impossible to say for sure, because Amazon isn’t telling). Authors who have never met but have exchanged e-words online have had their reviews on each others’ work removed. Fans who chat with the author too much (or at all?) also suffer.

There are a couple of things about this that deeply disturb me. First (and probably least), Amazon is gathering a lot of information about its authors and customers, in order to be able to identify these relationships. It’s unclear if this data-mining is done purely through Amazon’s site (and links from author pages to blogs, Twitter feeds, etc), or if it casts a wider net. Just how much is it watching us? (Though this is hardly surprising in light of what we know it can scrape off its customers’ Kindle devices.)

Secondly, this behaviour from Amazon means that the relationships that make being an author awesome are under fire. Indie authors, in particular, do well because of their interactions with their readers, with being accessible and visible. We enjoy great communication with our readers, and I think they enjoy being able to talk with us. (The same is true for traditionally-published authors, but usually to a lesser degree because they have more marketing options and support available than indie authors. However, it’s not my intention to under-value the impact on them; it affects all authors.)

Amazon is actively discouraging this. They’re punishing authors and readers alike for talking to each other by removing legitimate readers’ reviews. How is this good for sales? At the very least, Amazon is a business and should care about this. Fans who love a book or an author’s work are more likely to seek out contact with that author, so this could potentially remove an author’s most fervent support. Fans and readers don’t have a financial interest in the success of a book, so I don’t see how this violates the terms of service. Since when was it wrong to support something you love?

Similarly, this review ban-hammer affects authors who have had contact with each other. The author community is supportive and lacks the competitive viciousness of many industries. This is part of what I love about being a writer. We interact online. We swap notes and advice; we help promote each others’ work, because we know that a sale for other authors doesn’t mean less for me. And we know how important reviews are, so we like to leave them for the work we enjoy. Now being part of a community is wrong, too?

I’m not sure why Amazon thinks it’s a good idea to target these reviews. Direct review swaps between authors may be common, but does that mean that the reviews are any less honest? Of all the nefarious behaviour that has been identified around reviews, this has not been mentioned as being a problem, and it only covers a fraction of reviews by authors on another authors’ work.

Interestingly, the Amazon ban-hammer could lead to an unfortunate side effect: in order to be able to review books, reviewers are likely to create new (anonymous) accounts. So, instead of getting rid of anonymous sock-puppet accounts, Amazon is actually encouraging the practice, making it even harder to spot the ‘real fakes’ that caused the original problem. (This is an unverified prediction so far – has anyone seen this happening? Would you do it?)

Well done, Amazon, well done.

Thirdly (and possibly the saddest of all), they’re targetting the wrong reviews. There have been cases pointed out where obvious fakes are being left alone, while legitimate fan or fellow author reviews have been removed. On top of that, Amazon only seems to be targetting the positive reviews: viciously negative one-star reviews don’t seem to be affected. The revenge or false-negative review has been identified as being just as much of a problem as the false/fake-positives. So what’s being done about that? Nothing that we can see.

What this all means is that the quality of reviews isn’t being improved by the ban-hammer. It also means that the average star-count for books is being destabilised, because even if Amazon was removing the right (false-)positive reviews, it’s not removing the false-negatives.

So what does it all mean

It’s hard to know what to take away from all of this. Certainly, Amazon doesn’t understand the author-reader relationship, and it doesn’t understand the author community. It definitely doesn’t support these things. It’s one of those things that makes me very nervous about Amazon’s attempt to build a monopoly in the book industry, because if it succeeds, it won’t even have to pretend to care any more. It barely seems to give a crap now.

I don’t like that authors feel like they have to create anonymous accounts to leave reviews; that might be a natural reaction to the issue, but it’s a step in the wrong direction in my view. I hope this isn’t happening, or the sock puppet issue will only get harder to eradicate.

I think we should keep in mind that shoppers are pretty savvy these days, when it comes to reviews. They’ll ignore the obvious too-good and too-bad reviews, and look for the ‘honest’ ones. So why should Amazon headache about it like this? It feels like a badly-aimed reflex in response to the reported problems and abuse.

What do I think Amazon should do? Whatever algorithms it is currently using are clearly faulty. Stop it. There is a facility to report something that looks dodgy, so let people report dodgy reviews and investigate them properly when they’re identified. Identify the markers of actual fake reviews and step on them. Make the criteria clearer for what qualifies as a ‘close personal relationship’ (the mechanism they use to assess it might be intellectual property, but the basic criteria it has to meet should not be). Stop using lazy measures that annoys everyone involved (including our customers!).

One thing all of this has meant is that there is now a petition accepting signatures to appeal against the review ban-hammer at Amazon. As of this posting, it has over 15,000 signatures. Indie author Jas T Ward explained the petition’s intent in an email to Amazon CEO Jeff Bezos just a few days ago.

If you feel strongly about this issue, I suggest you sign the petition. If you have had a review wrongly removed from Amazon, shout out about it! Affecting the direction of a behemoth like Amazon might be a difficult task, but that doesn’t mean we shouldn’t try.

I am going to be writing up some reviews in the near future (less writing means more reading time for me!), so I’ll be curious to see if the ban-hammer swings in my direction. I’ll be posting the reviews in multiple places (including here), so they won’t be able to remove all of them!