What have the 2015 UK general election and digital analytics got in common? Perhaps not much but we can still learn something about user intent from the election.
During the lead up to the UK general election the polls were calling a very close result between the two leading parties, Conservative and Labour. These are polls conducted online and over the phone and they seek out voter intent in forthcoming elections and in some cases they also ask about voting behaviour in previous elections.
In the 2015 UK general election the polls got it badly wrong by overestimating Labour’s performance and underestimating the Conservative performance.
What’s interesting is that on election night the exit polls were accurately reflecting the final outcome and many, including Lord Ashdown (a Liberal Democrat grandee), were surprised and in his case rather incredulous.
We all know the result.
What this seems to suggest is that if you ask somebody about future intent the answer they give will quite likely be less accurate than any answer they give in relation to past activity. That makes total sense, people can change their minds but they can’t change history.
From politics to digital analytics…
Translated into the online world, if clickstream data tells you what is going on and voice of visitor data tells you why it’s happening, it’s worth remembering that any answers in relation to future intent could be misleading.
Professor Daniel Kahneman discusses two types of thinking in his book “Thinking Fast and Slow”:
- System One which is emotional, instinctive and fast
- System Two which is methodical, logical and slow.
When answering questions in a controlled user testing environment people might deploy System One as they feel under pressure to perform for the moderator (or give an immediate answer to the pollster’s question), but in the cold light of day when users are alone and weighing up a purchase they may well think and behave in a very different way, deploying System Two .
So in considering a new element of functionality to improve an area of the on-site customer journey digital analysts, UX people and the like should be wary of how they couch their questions when conducting qual research.
It may help to consider if the new element of the user experience is tailored more towards evoking an emotional response or driving a logical decision.