You are going to invite a friend to a restaurant and want to know if they serve good food there. Or you intend to rent a flat and need to discover whether the estate broker is reliable. The easiest way to find out is to google the name of the restaurant or the broker. Most likely there are enough reviews of other customers to give you a basic idea. In fact, you may be well looking at fake construction of reality rather than facts. According to this study, the server Yelp has to filter 16 % of the comments concerning restaurant, since they may be fake and pre-ordered. Undetectable part of fake reviews still goes unfiltered, also on other servers like TripAdvisor or Google Maps. This April, Amazon sued several companies that had been writing fake customer reviews on Amazon sites for money. Such top of an iceberg indicates that discussion fora can turn to battlefield of secret commercial campaigns. Small vocabulary below explains most important notions of customer discussion manipulation.
Apostle of Myth – a person who has read fake review online and blindly keeps forwarding its message (a myth) on other fora. Apostles are people who like to chat, contribute to discussions or look clever. They unknowingly serve the myth disseminators.
Corrupted Thread – A line of discussion that is manipulated and/or contains large number of fake comments. Those comments usually come from the same IP address and use similar language.
Countermeasures – For companies that are attacked by hired or biased reviewers, it is hard to respond. The best defence is to announce an explanation that there are fake reviews around and to let simple facts prove why they are false. After all, negative advertisement is still advertisement. Professional companies offering to redress the situation like “reputationmanagementconsultants.com” may in fact use the same tactic of manipulation.
Disseminator of Myth – person who pretends to be a customer and writes an artificial review. Myth disseminators may be professional and claim to be “objective” like buyreviewsnow.com. Or they may be simply part of a business PR team.
IP Address – unique identification of every computer, visible to the administrator of websites.
Myth – to succeed, a fake review must contain a catchy and coherent message or idea, a myth. E.g. “in the XY restaurant, they add horse and rat meat to the burgers. My uncle is a health inspector and he discovered it.” If the myth is good, apostles will keep spreading and exaggerating it – “(I heard that) they serve horses at XY!”
Offence – legal qualification of fake reviewing depends on country and type of review. False positive review may be unfair business practice, while negative review can be a libel or slander, entitling the damaged party to monetary compensation. If the allegation in the fake review is serious, its author may have committed the crime of defamation. Identifying the perpetrator and gathering evidence in digital freedom-of-speech era is rather difficult.
Recognition – How to recognise false review or corrupted thread? First indicator is the increased amount of superlatives without any further description: “absolutely breath-taking and incredibly wonderful book that changed my life”. Secondly false reviews often contain appeal to action: “Do not ever go to dine there!” or “You must buy it now!” Thirdly there are often references to what others do or say: “My grandma bought this blender and was totally overwhelmed” or “Police already investigates this company”. Moreover, the myths in the review may seem illogical and exaggerated even by punctuation and special signs like exclamation points or smileys. Corrupted threads also often contain lot of comments from the same IP address. Unfortunately, a review with all those clues may still be genuine and a review without any may be a fake.
Responsibility – In Delfi judgement, European Court on Human Rights recently confirmed that the owner of a server is responsible for offensive content of the discussion. If the server administrator is not able to remove the offensive content within a reasonable time, the owner might be ordered to compensate the damages caused by the comments. However this does not solve positive fake reviews or arising necessity to examine tens of thousands of reviews uploaded every day.