Tuesday 4th July 2017
Earlier this year, Tom Woodnutt and Hugh Carling presented a paper on the new science of Behavioural Recruitment at the Ilex conference on the future of market research.
The underlying premis is that you need to get the right 'bums on seats' for the research results to be valid - and therefore that robust and accurate recruitment is the foundation of good market research.
However, as we have seen from comments on the egroup, the rise of the 'professional respondent' is challenging this - are they telling us the truth about their behaviour or prefences (or are they just after the incentive?) and does the fact that they frequently take part in research mean that they are truely representative of a typical consumer. This is not a new problem - many years ago as a young moderator I picked up an 'invitation' card left behind by a respondent. On the card was written 'For this session you are a customer of XYZ bank' - which unsurprisingly gave me real concerns about whether or not they were!
Having authentic respondents who are representative of the target market in question is crucial for the industry - and for the perceived validity of the insights gained.
So is behavioural recruitment the answer? According to Tom and Hugh, using data available on social media we are able to identify potential respondents according to their actual behaviour - where they have been, what they think and how they react rather than relying on what they say they've done (which we know is subject to post rationalisation and bias). Harnessing the vast amounts of data that platforms like Facebook have about its users means that we can find people who genuinely match specfic criteria and are engaged in the subject.
Whilst I can see limitations in using this approach, I can also see some real value in using social media to help power recruitment - and I am sure that the technology and how we harness it will develop even further in the next few years. Tom and Hugh's paper is interesting and thought provoking and definately worth a read.