Eurostat, the umbrella organisation for the statistical institutes of the member states of the European Union, has just published some worrying results about the trust Europeans place in social media.
It has to be said that the champions of mistrust are the French, with less than 20% of those surveyed considering them to be "reliable". However, Germany, the UK and Spain are scarcely more optimistic, with rates of between 20 and 30%: only in Poland does confidence prevail.
More than the regional differences, what is striking is their evolution over time: a few years ago, social media - with the first blogs in particular - appeared to be the champions of a new freedom of speech, a real counterweight to the established media.
It was the time when Carlo Revelli and Joël de Rosnay founded Agoravoxthe 100% is a "citizen and participative 100% media". Today, the site is home to a large number of revisionist and negationist papers, with all the nauseating implications that can entail.
Since those heroic and particularly utopian days - when Web 2.0 was going to help us build a better world - many agitators have realised how easy it is to influence crowds by distilling a multitude of untruths.
The phenomenon extends far beyond France's borders: we know the role played by social networks in the American presidential election, where it now even seems established that Russian hackers played an active part in the defeat of the Democratic candidate.
During the 2005 referendum on the European Constitutional Treaty, an analysis of what people were saying on blogs and forums very quickly showed that the victory of the 'yes' vote was not certain, contrary to what the majority of opinion polls were predicting.
Now that these same polls are struggling - Brexit, the election of Donald Trump - social media are not helping to refine our vision of society; on the contrary, they are disrupting it - at least if you read them at face value, without decoding rumours and manipulation.
In consumer research, where the stakes are more modest, it is easy to cross-reference declarative and behavioural data to better identify buyers and users. In the context of political research, where time pressure plays a very important role, the compasses will certainly be going wild in the coming months.
Be that as it may, recent developments in social media represent a new reality that needs to be taken into account: just as we can't be satisfied with a first-hand reading of the statements made by the people we are interviewing, it is equally important to identify, behind the majority of opinions posted, the more or less malicious strategies of influence.