Categories
Uncategorized

Portuguese consumer organisation Deco warns: some Amazon, Booking.com, or TripAdvisor reviews may be false

The site Amazon has the largest number of suspects Comments (8.4%) in a study involving Deco and three other associations of European consumers. Second is TripAdvisor (6.2% of the statements are doubtful) and, third, Booking (2.1%).

Between June and August of this year, they analyzed 6.3 million opinions for 46,500 products and hotels, and concluded that “some of the opinions contained on the sites of the three platforms may be false.”

The results indicate that 60.5 thousand reviews on Booking.com raise suspicion as to its veracity, as do 14.6 thousand reviews published on Amazon and 10.7 thousand on TripAdvisor.

Associations explain that those online service providers have developed “some tools and strategies to minimize suspicious opinions,” but not only did they not solve the problem, but they don’t seem concerned, consumer associations say.

“The results of our study would not be very serious if the platforms we looked at were concerned with the false ratings that are published. But, on the contrary, the poor control and permissiveness of the criteria for posting opinions mirrors the careless and timid way try to control this phenomenon, which is very far from the standards required to defend the interests of consumers “.

Consumer representatives believe that there should be a validated register of users making comments; the obligation to make a purchase to post an opinion; prior checking of user trustworthiness and comment; the creation of mechanisms that detect both sudden increases in comments and user verification to reduce the possibility of robot intervention.

It is the practices they want to see applied by Amazon, Booking.com, and TripAdvisor, who will present the results of the study.

What was analyzed?

The technicians analyzed the comments and who made them, in 11 parameters, with different levels of importance and weights. In particular, they verified the existence of repeated sentences in different evaluations, the concentration of opinions in short periods, the deleted comments, unverified purchases and the amount of opinions of the same user.

They also took into account the profile of the evaluators: whether they only gave one opinion or made several comments on different products, for example. They checked if the reviewers always gave very positive opinions and if there were comments deleted by the platform or regarding unverified purchases.

The analysis was done in two periods to check for deleted comments and / or deleted users from July to August.

“When a product failed in many ways, significantly altering the final score, we thought it was affected by opinions that were suspected to be false,” write the technicians.

What is suspicious?

Understanding false comments is not easy, but associations leave some warning signs:

– Verify who is responsible for the transaction, if you are a vendor outside the platform he may be tempted to illicitly promote the products.

– A product has many excellent ratings (five stars) and at the same time many negative reviews (one star).

– Older opinions all have very high scores compared to the latest and coincide on certain dates.

– Give preference to verified comments made by those who actually purchased the product

– See other comments by the same author. You have purchased more than one product from different brands and rated them all with five stars; bought several products of the same type; never gave an opinion on an article from a recognized brand?

– Opinion includes many photos or even a video. “It’s unusual for someone who is satisfied with a product or service to put so much effort into promoting it,” they warn.