Elections or reconciling with the polls...
For two months, nothing will happen in France.
Our country will be living to the rhythm of the Presidential election, its scandals and the pace of the opinion polls: "the institutes had better behave themselves", we'll say, because if they repeat the Brexit, the elections of Donal Trump, François Fillon and Benoit Hamon in the primaries, their credibility will be gone forever.
It's easy to forget that polling is not an exact science and that the material studied is constantly changing: only quantum physicists are in a similar position. In this respect, this election has never so clearly demonstrated the need to use a qualitative approach to analysing polls. Because today, beyond the values in %, we can only be sure of 3 things: the French are angry; they want political lyricism; they are oscillating between heart and reason (we'll leave you to interpret!). But there's no doubt that after the elections, they (= the journalists) will be firing on all cylinders at the polls, which they (= the journalists) will have abused for a year!
Especially as certain observers will be stepping in to disrupt the debate and try to get their own way on the theme: quantitative research is out, all you have to do is read what contributors are saying on social media.
There's so much being said on the social web, and so many ways of gathering conversations, that with a bit of luck there will always be a winner: the one who gets a better result than the 'old' institutes.
The self-proclaimed winner is a Canadian who announced the victories of Donald Trump or François Fillon, where they were not expected ... but not Brexit: it doesn't work every time.
At the time of the Right and Centre primary, the Canadian Web specialist gave François Fillon a score of ... 22% the day before the election, i.e. half his actual score, just ahead of Nicolas Sarkozy. Web listening is obviously not an exact science.
With the right storytelling, you can get a lot out of figures!
If you want to be the talk of the town at the next Football World Cup, bet your friends that in the final there will be two players on the pitch born on the same day!
You're not taking much of a risk: in a group of 23 (= the number of players in a selection), the probability of two people having the same birthday is precisely 50.73%; and as there are two teams playing against each other ...
Finally, it's worth remembering that figures don't always read the same way in politics as they do in marketing.
If during an election a candidate obtains 51% of the votes, he is proclaimed elected against an opponent who only obtains 49% - and there is no margin for error: there is a winner and a loser.
But if, in a survey of 1,000 people representative of your target audience, 51% of consumers say they think your product performs well, watch out! 51% or 49%, it's the same thing: you're in the error zone.
So, if your product only satisfies half of your potential customers, the only reasonable decision to make is to redesign it, because that's a long way from real success.
