Interesting New York Times article

... scary stuff.

Mr. Liu estimates that about one-third of all consumer reviews on the Internet are fake. Yet it is all but impossible to tell when reviews were written by the marketers or retailers (or by the authors themselves under pseudonyms), by customers (who might get a deal from a merchant for giving a good score) or by a hired third-party service.

Click to expand...

Hopefully the FTC will put him and his company out of biz. He's doing the public a tremendous disservice. Honestly disgusting.

Ugh. I really hope it doesn't catch on.

Yet it is all but impossible to tell when reviews were written by the marketers or retailers (or by the authors themselves under pseudonyms), by customers (who might get a deal from a merchant for giving a good score) or by a hired third-party service.

Click to expand...

It may be difficult to tell which exact category the fake review falls into, but the fakeness of most fake review is obvious to any but the most most oblivious reader.

(A question for Mr. Liu might be, "If you can't discern which reviews are fake how do you know that 1/3 of all reviews are fake?")

James D. Macdonald said:

It may be difficult to tell which exact category the fake review falls into, ...

Click to expand...

In some cases, but I think it's pretty obvious that any (even slightly) negative review of my book is a fake.

Ken said:

Hopefully the FTC will put him and his company out of biz. He's doing the public a tremendous disservice. Honestly disgusting.

Click to expand...

First, he has to be breaking a law, and I'm not sure he is. Neither am I sure that laws regulating book reviewing are a good idea. Reputation of the reviewer is about the only way of validating a review.

On the other hand, faked reviews or reviews by people who haven't read the book are often easily spotted by the proliferation of superlatives as well as their insubstantive nature. Often, they don't mention any of the events in the story, or name any of the characters.

But there's a spectrum of honesty to reviews, beginning with the honest critical review, to reviews by friends who are anxious to support their friend's book, to reviews that are compensated, to totally bent reviews by people who don't care about or read the book. Where to draw the line?

johnhallow said:

Ugh. I really hope it doesn't catch on.

Click to expand...

I was reading about paid reviewers ten years ago, and I'll bet it's been going on since Gutenberg printed his first Bible.

The insidious thing about buying reviews is that the influence of those reviews on readers isn't the point of the exercise.

The issue is that large numbers of reviews and high star rankings figure (or used to figure) into Amazon's mechanism for deciding which books to recommend. If your book is featured prominently on Amazon, a lot of people see your book. Seeing your book is a precondition to buying it, and getting people to see your book is a major hurdle for authors to get over in an Amazon ecosystem that includes over 7 million books.

Amazon's various systems are designed to automatically recognize trends and memes and whatever else generates heat behind a book, and to get that book in front of its audience when it starts blowing up. It will move books into "recommended for you" lists on the front page, and it will add the books into the "customers who bought this also bought" lists more easily. It will even e-mail customers about the book.

Amazon is designed to identify whatever is popular or becoming popular and makes it more popular by essentially providing a huge surge of free advertising through prominent positioning on Amazon's site. So a very good way to actually become popular is to trick Amazon's computers into thinking you're popular already.

John Locke was able to do that because he spent a lot of money to contract for paid reviews in huge volume when nobody else was doing that. Now, fake reviews are very cheap, but much less valuable because they will no longer trigger Amazon to do all the things that Locke's fake reviews did for his book.

At the time, Amazon wasn't looking for authors to orchestrate hundreds of fake reviews, so Locke's fake review activity looked like an emerging bestseller to the algorithms, and they started promoting his books. Even at the $65 per review he evidently paid, according to the article, this was probably much less expensive per-pageview or per-click than Google ads or Facebook ads.

There are lots of other similar shenanigans that people have tried. For example: paypalling hundreds of people $2 each in exchange for buying a $0.99 e-book on Amazon at a specific time to engineer a ranking spike.

And Jeremy Duns, who outed Locke, keeps getting his Twitter account suspended. No doubt this is not the work of a Get Jeremy Duns Suspended From Twitter Mafia.

Now, fake reviews are very cheap, but much less valuable because they will no longer trigger Amazon to do all the things that Locke's fake reviews did for his book.

Click to expand...

Very interesting. That does sound like the crux of the problem. So, has Amazon changed its popularity algorithms in response to this? (Read the article yesterday, but if it mentioned that, I can't remember.)

I stopped reading Amazon reviews when I realized that, for every self-published book I looked up after someone sent it to me, I would find either no reviews or a bunch of generic five-star reviews that told me nothing. (These were obscure books, not best sellers.) For my personal reading, I use Goodreads, where I usually find some three- or four-star reviews with actual information in them. I've learned to ignore one- and five-star reviews — genuine or not, they just don't help me when I'm on the fence about buying something.

I admit, I was drooling as I read how much money Rutherford made from fake reviews. I receive dozens of self-published books for review each month. I cull them and read and review a few. Though I get a salary, it doesn't really cover the time I put into this. Writing fake reviews would be much easier and, apparently, more lucrative.

But I have these inconvenient scruples and a general hatred of everything that detracts from the credibility of honest reviewers, so it ain't gonna happen.

Fuchsia Groan said:

Very interesting. That does sound like the crux of the problem. So, has Amazon changed its popularity algorithms in response to this? (Read the article yesterday, but if it mentioned that, I can't remember.)

Click to expand...

Amazon did not reply to the NYT's request for a comment or an interview, and Amazon's mechanisms are not discussed in the article.

The NYT writer, presumes, like many people, do that the fake reviews function as persuasive testimonials to potential readers. I seriously doubt this; just about every book that hasn't been the subject of some kind of specific backlash has a rating between 3.5 and 4.5 stars on places like Amazon and Goodreads. Very obscure books often have near-perfect ratings because only friends of the author have reviewed them. And they stay obscure, despite those high star ratings, because the content of the reviews doesn't matter.

Generally, once real readers start looking at books, the best reviews tend to get up-voted and the few reviews you see on the front of the page are decent, with the spammier ones buried.

djf881 said:

The NYT writer, presumes, like many people, do that the fake reviews function as persuasive testimonials to potential readers. I seriously doubt this;

Click to expand...

I have an intelligent friend who used to believe the reviews on Amazon until she showed me a book she'd bought based on all the 5 star reviews. I read it and told her I thought it was really bad writing and why. And she said she could see my points and wondered why it had gotten such good reviews then? All 5 star, nothing negative at all?

She was surprised when I said many of the good reviews are written by friends and family and aren't true reviews at all ... she just hadn't thought of people doing that. She was naive and trusted the reviewers.

I'm not saying all good reviews are fake, not at all. I've written good reviews on books I really liked. But if a book only has 5 star reviews, all gushing and cloying, then I run away.

As to the article - man, that guy was pulling in $28k a month for writing reviews! More than almost any novelist, I assume. Glad that operation was closed down. But there will be another one along, probably.

This is just one more reason why I want the credibility of conventional publication (all digits crossed!) and would only e-publish as a last resort.

You can find book bloggers who still review self-published works. They'll generally have guidelines. Follow those guidelines to the letter.

Meanwhile, so far as I'm aware, no one has figured out how to game Published By Random House, Reviewed in the Times. Other than by writing an excellent book, that is.

juniper said:

I have an intelligent friend who used to believe the reviews on Amazon until she showed me a book she'd bought based on all the 5 star reviews. I read it and told her I thought it was really bad writing and why. And she said she could see my points and wondered why it had gotten such good reviews then? All 5 star, nothing negative at all?

She was surprised when I said many of the good reviews are written by friends and family and aren't true reviews at all ... she just hadn't thought of people doing that. She was naive and trusted the reviewers.

I'm not saying all good reviews are fake, not at all. I've written good reviews on books I really liked. But if a book only has 5 star reviews, all gushing and cloying, then I run away.

As to the article - man, that guy was pulling in $28k a month for writing reviews! More than almost any novelist, I assume. Glad that operation was closed down. But there will be another one along, probably.

Click to expand...

There's nothing especially sinister about the fact that reviews tend to be overwhelmingly positive. People tend not to buy books they don't expect to like. I may have my own assessment of books where 3 stars is the average book I read, but the average book I read is a very good book, because I only read good books. I think, when compared to the larger universe of books I don't bother with, the books I read are almost all four or five star titles.

The review culture also skews toward positive reviews. Anything less than 4 stars is basically a pan, so most reviewers give things 4 or 5, unless there's something really disappointing about the book. A community norm where five star reviews were reserved for very special books would probably be more useful, but that isn't the norm.

I read between two and four books a month, and with a few exceptions, I'd rate everything I've read in the last two years four or five stars.

Another said:

Is this the place to ask how to get honest reviews of self-published work? I realize there are threads on marketing published work, but perhaps self-published material is a special category. Any relevant threads on WC?

Click to expand...

The trouble is that there are going to be over 200,000 self-published books this year, and there's really nobody out there with a professional interest in identifying the relatively good ones.

On top of that, traditional publishers are pushing into the low-priced market with Kindle Daily Deals and various sales on backlist titles. This is the flip side to self-publishing that Konrath and his acolytes don't tell you about. It's very hard to get people to read a self-published book.

If you're traditionally published, you'll probably be reviewed in at least one or two of the trades, and good reviews there can really draw attention to you. Many of the best book bloggers; librarians and other credible reviewers stick to traditionally published titles. The top-ranked Amazon book reviewers also stick mostly to traditional titles.

That's why almost everyone who can publish traditionally goes that route.

Transatlantic said:

This is just one more reason why I want the credibility of conventional publication (all digits crossed!) and would only e-publish as a last resort.

Click to expand...

I generally agree, but ... In the past year, I have bought the only two self-published books I've found outstanding. I sampled both because of reviews (a blogger in one case, several Goodreads reviewers in the other). When a reviewer is highly articulate, seems intelligent, compares the book to other books I like, does not give positive reviews to everything he or she reads, and cites specific examples to support the book's worthiness, then I pay attention.

Well, at least enough to read a sample. If it is a good book, that will do the rest.

One Example?

James D. Macdonald said:

You can find book bloggers who still review self-published works. They'll generally have guidelines. Follow those guidelines to the letter.

Meanwhile, so far as I'm aware, no one has figured out how to game Published By Random House, Reviewed in the Times. Other than by writing an excellent book, that is.

Click to expand...

Yup, I get that. But how about one example of a blogger-reviewer for self-published literary fiction? Call it a love story.

Reviewer Market?

djf881 said:

The trouble is that there are going to be over 200,000 self-published books this year, and there's really nobody out there with a professional interest in identifying the relatively good ones...

Click to expand...

Sounds like a market opportunity for a good, trusted reviewer or two or three to steer readers through the trash to the good stuff. No?

Another said:

Sounds like a market opportunity for a good, trusted reviewer or two or three to steer readers through the trash to the good stuff. No?

Click to expand...

Check out the last several pages of Authors Should Really Stop Telling Readers How to Give Reviews
http://www.absolutewrite.com/forums/showthread.php?t=230972
and you'll see how that's been working out.

The guy who was selling paid-for reviews in the Times article justified himself by saying that later reviews by actual readers giving lower rankings would balance out and correct for the paid five-star reviews.

But we've seen that giving anything less than a five-star review is called "bullying" and opens the reviewer to personal attacks and harassment both on and off line.

Amazon keeps its exact recommendation algorithms a closely-guarded secret and is constantly tweaking them to keep people from gaming the system.

For Sale: Answers to the final exam in Ethics 101. Guaranteed to get that A in the Ethics class on your permanent transcript. Cash only.

Transatlantic said:

This is just one more reason why I want the credibility of conventional publication (all digits crossed!) and would only e-publish as a last resort.

Click to expand...

There is a definite credibility to convential publication and I wish you luck. You'll probably find that reviews are still important to future sales, though. Those are tough to get sometimes.

My small press publisher sent out review requests in re: my paranormal romance to over one hundred sites. No interest. I knocked on doors until two reviewers finally read the material. Now that I self-pub, I've put out six novellas and received one Amazon review.

Maybe I should have just bought some reviews. Less time + guaranteed results!

AnneGlynn said:

Maybe I should have just bought some reviews. Less time + guaranteed results!

Click to expand...

That's why some guys go to hookers.

Aside from getting diseases, getting robbed, getting blackmailed, losing the respect of your friends, and losing respect for yourself, what's the downside?

Well, knowing that you were so desperate that you had to pay for it....

So anyway, Satan appears to a writer and says, "I can give you a thousand 5-star reviews."

The writer says, "What do I have to do?"

Satan says, "You have to sell me your soul and turn over your wife and kids to me, so that I can torture them throughout eternity."

The writer says, "Okay, but what's the catch?"