The latest studies set a 23% the number of SMEs that are giving way to the world of social networks and websites of recommendations, although this percentage grows meteorically every day and we hope to see it take off exponentially in the coming times to reach the almost totality of Spanish companies. However, we are still in full swing and there are many SMEs that still offer strong resistance to the jump to networks and the world of opinions and recommendations made through the Internet.
Precisely one of the most frequent fears injected is the inability to assimilate or face opinions with no more negative, but fraudulent: unfair competition, angry workers, disgruntled customers and a long list of agents difficult to identify that clearly try to harm the business. In this same bag we find what in the industry is known as “positive spam”, a name under which we include companies that generate hundreds of false profiles with the sole purpose of speaking positively about their own business. In short, every day we find new cases of users and companies that “skip the law” in one way or another with the obvious aim of distorting the image of certain businesses.
This fact requires solutions, improvements and formulas that reduce as much as possible the cases of spam and fraud, protecting not only the image and interests of the businesses, but fundamentally the user that consumes information. From this endeavor, we can extract some rules that should be made available to all recommendation websites in order to offer certain guarantees, both to companies and users, that the opinion they read can be filtered according to the prism of truthfulness.
Lean on technology. Although a priori it seems difficult to detect patterns that lead us to classify an opinion as “truthful” or “fraudulent”, little by little we have found automatic filters and tools that help us determine the quality of an opinion with less margin of error each time , taking into account variables such as the origin of the opinion, global activity of the user, time of publication of their opinions …. We are sure that over the next months we can refine the process even more and add even more consistent filters .
Add the human component. In the end, separating the grain from the straw is a complicated task. For many technological filters that you put, it will always be essential to have one or several people on the team examining the automatic processes, reviewing the suspicious opinions, making the necessary arrangements and verifications and making decisions on how to act.
Communicate the conclusions of each case. The human component of the previous point is not exhausted in the more “bureaucratic” functions. It is very important to have a human team capable of listening to users and companies, review each case in detail, follow up and always, always give an answer to those involved in each case.
Evangelize your companies I: do not fight back. When the company encounters an opinion that it considers fraudulent, its measures to channel it often take the form of false profiles to counteract the “attack”. It is very important to teach companies that good management is to avoid measures of this type: on the contrary, it is more like throwing stones against their own roof. The false profiles are easily recognizable by any user, “sing” everywhere and do nothing but aggravate the bad image of the company and the credibility it generates among users.
Evangelize your businesses II: encourage the responses of the owners. Offering your companies mechanisms to officially respond to the criticisms they receive, good or bad, and teach them the best way to answer and deal with them, will always result in a clean and transparent image of the company, however bad or fraudulent it may be. opinion.
Do not be afraid to block or delete opinions, false profiles, problematic companies … We are often scrupulously careful to eliminate any content from the web to avoid falling into the swampy terrain of what many call censorship, but when a company or a user they show bad behavior, do not contribute anything to the conversation or even damage the credibility of the web itself with its actions before the rest of the community of users and companies … why allow it to continue acting? Applying restrictive measures very rarely protect most of your community.
Do not allow unfounded opinions. It is often difficult to discern if an opinion is based on real knowledge of a site or business or comes from third-party references. Even so, many users clearly state that they do not know the place they criticize first-hand. In that case, the action to be performed should be clear: out!
Invite your users to be “real”. Of course, on the Internet each one decides the identity with which he decides to operate; However, the “anonymity” of the users is one of the biggest fears of the companies and one of the tricks they most frequently use to discredit the opinions they receive. Therefore, it is advisable to encourage users to use their name and surnames to give credibility to their opinions and generate trust among other users and among the companies they think about.
Have well-defined criteria and firmness in its application. Often many webs give way too lightly to the pressures of companies (often paying customers) to withdraw opinions that go against them. We think it is important not to give in to a situation like this, having well-defined criteria to see when an opinion (good or bad) is true and when it is fraudulent and, depending on the decision made, communicate it to the company.
Investigate, investigate, investigate … The fight against fraudulent opinions and spam does not end with a few actions; It requires a constant effort to investigate and improve with new technical filters, train and grow the team that manages these cases, try to get ahead of the “tricks” of the trolls, etc. From our point of view, it is important to allocate a good amount of resources to this topic: the maintenance of the quality and credibility of a web of recommendations is undoubtedly a basic pillar for its smooth running.