It was “Brexit plus plus” wasn’t it? Trump was right. He confounded the media and the pollsters, just as the Brexit vote did. So why did the pollsters get it wrong?
There are two different factors at play here:
- Honestly answering the wrong question
- Dishonestly answering the right question
The survey below which asks ‘What won the election for Donald Trump’ illustrates how this can happen.
Look at Q2. Like most people I ticked ‘because Trumps message resounded with people who weren’t accurately polled’. I don’t actually believe that, it was just the closest option to my views.
Yes his message must have resounded with people – they voted, he won.
But is it because people weren’t accurately polled? No, I don’t think so.
Accurate polling means sampling a representative cross section of the population and asking them which way they are going to vote.
What can you do if people lie? It wasn’t politically correct to say you’d vote for Trump. As Paul McNamee wrote in the Big Issue: ‘his distasteful dog-whistle remarks rose like warts during the campaign’.
Not telling the truth and answering the wrong question honestly can happen all the time in market research if you’re not careful.
Here’s an example of how to overcome this problem.
I once ran a series of focus groups for an education client where, after the ‘official’ focus group was over, we stood around outside and chatted informally about the topics we’d discussed. In these informal discussions, we gathered a completely different set of answers – which were the answers we used to correctly develop the product.
Lessons learned? It was silly to ask the question “Will you vote for Trump?’ given how un PC he was, so the pollsters got a silly answer.
You have to ask different questions in different ways to get the right answer.
