Amazon Poor Shill ReviewsThere is no secret that online reviews are an immensely powerful piece of the conversion jigsaw.

Who doesn’t go searching for reviews of products / services that you are considering buying? I know that is one of my first ports of call when launching off on a purchase journey (unless I am feeling a bit trigger happy!).

I noticed something on Amazon yesterday that I hadn’t seen – the description of a book ended with the following disclaimer:

Note: This book along with others has had some poor shill reviews posted on it in order to manipulate sales.”

I found this interesting for a number of reasons:

  • I had no idea what a ‘poor shill review’ was
    • As an aside, a quick Google search on the warning message led to me seeing the exact book in position 5 of Google’s organic results – who says that results aren’t personalised? (I was using a different browser window and wasn’t logged in on either…)
  • There was no indication whether these reviews had been removed or whether Amazon believed that some of the 22 reviews on display were effectively fake
  • It shows how big the problem of fake reviews must be

Hats off to Amazon for tackling it head on rather than pretending it doesn’t happen (although I think that it warrants a bit more of an explanation?) but it is a reminder that you can’t always trust what you read.

How does this leave Google and the notorious ‘social signals’? If a behemoth such as Amazon is coming out and saying that not all reviews are genuine and have been created simply to influence buying decisions, does it not raise questions about using social signals as a key ranking factor?

The simple truth is that people will try to game the system, whatever the system may be, and I am not for a second suggesting that looking at social signals is not a sensible indicator of quality but I was surprised to see the warning on Amazon. Maybe I have been blind to it before, but is this a relatively new feature?

The beauty of social signals is that the community at large can endorse (or destroy?) content. By asking the masses, you are reducing the potency of a biased review and you would hope that a few rotten apples will not spoil the barrel.

We have to rely on the algorithms making sense of all the noise, but a truly great product should always attract positive feedback and it will be hard to skew the overall impression of that product if there is a groundswell of support for it.

It is rare to read reviews these days without finding some negativity and I personally find the odd negative review to be strangely reassuring as it would suggest that the reviews have not been fabricated entirely by the brand in question. People are always much more likely to moan than they are to heap praise on something (in public) so I always expect to find negative reviews if I look hard enough.

All online users should be treated with respect – they are not idiots and do not need telling that not everything they read online is true. That said, I think Amazon is strong to quietly remind its users that not all reviews are genuine.

Over to you Google – how do you propose to deal with the trolls and social signals that are not entirely genuine?