That’s decided, I’ll book the hotel that received the best comments on booking.com. This is the way to successful holidays! More than 70% internet users look at other consumers’ reviews before they buy, but only 25% find them reliable. First, common sense tells us not to trust blindly other consumers, and second, the increasing number of fake reviews justifies this suspicion. About half of the reviews would be fake according to the DGCCRF (French general office for trade competition, consumption and repression of fraud).
Andreas Munzel builds his Recherche et Application en Marketing article on this fact. He shows how we can help internet users identifying fake positive reviews to make more reliable choices.
Three experiments respectively conducted with 197, 211, and 141 US consumers (average age 30) show that a website can use two strategies to appear reliable, to receive favourable evaluations, and then increase the willingness to buy.
- It can use a 3rd parry-awarded label (here from Consumer reports) stating the website has procedures to detect fake reviews
- It can simply state for itslef its ability to identify and remove fake reviews.
The 3rd party label strategy proves more efficient. The website appears to be more credible and triggers a higher willingness to buy. It works even better when internet users have previously been warned about potential fake reviews and the difficulty to identify them. Older consumers and those considering themselves as experts (in reviews) are more likely to follow this pattern.
OK, but how can one do?
In relation with the efficiency of the 3rd party label, the author advises setting up the (non compulsory) AFNOR norm. Since July 2013, AFNOR (French national organization for standardization) suggests a list of good practices to websites willing to give credentials to internet users about their reviews. He also recommends using labels such as Fia-net, Avis verifies, etc. Only 3 out of 10 websites have certified reviews at the moment.
In addition, websites must continue their efforts to improve the reliability of algorithms in charge of detecting fake reviews based on the words used, the repetition, the length of the comments… but also on the integration of “contextual factors such as the consistency of the review, or the divulgation of information about the author of the review” that are less likely to be manipulated by the pros of fake reviews.
Fake reviews trigger a generalised suspicion, which could backlash against the websites’ owners. Ironically, some owners are at the origin of the production of fake reviews as they buy them from professional companies. When one’s hoisted by his own petard?
Munzel A. (2016), Malicious practice of fake reviews: Experimental insight into the potential of contextual indicators in assisting consumers to detect deceptive opinion spam, Research and Applications in Marketing, 30(4), 24-50.