"In June last year, soon after the general election, acting leader of the Labour party, Harriet Harman, got in touch. She wanted to commission voter research to help to understand why Labour lost. I, along with colleagues, Ben Shimshon and Cordelia Hay, undertook that work and conducted focus groups in places with particularly disappointing results (Croydon, Edinburgh, Glasgow, Nuneaton, Watford). Our report was intended to sit alongside Margaret Beckett’s review, finally released this week. However, it remains unpublished."
The big political row at the moment is over how the Labour Party fared during the last General Election. For the record, I did tell a local activist Labour would lose - and why- that was six months before the election date. ITV's new political editor young Peston trumpeted that he had obtained the original report. Imagine my utter disappointment to find the report just related to four focus groups.
I say disappointing as it was in 1973 I helped introduce to the UK FOCUS GROUPS! I conducted some of the first focus groups in the UK based on a model imported by the advertising agency Foote Cone & Belding (FCB) from Madison Avenue, by way of Wall Street in the USA. FCB was set up in Baker Street London and yes the décor and staff straight out of MAD MEN. Wonderful stuff! Legendary Lunches too ...
My lasting criticism of how the focus group methodology has been hijacked and torn apart by political policy wonks has been commented on and 21st Century ,so called political research is simply based on the simple fact that they are not doing it right and therefore the results cannot be relied upon. THEY ARE GETTING IT ALL WRONG. This is why.
Focus Groups dictate that the end purpose isn't revealed to the panel. Panel selection is based on deep and careful socioeconomic profiling, to reflect the target audience required. The 'book' or script for the focus group ie questions asked, situations or products handled. is prepared and pre-tested. Two or three versions are prepared.
The hidden (it used to be one way glass) control group of colleagues - these days should see the output of the group via a video link - would then feed in additional materials depending on the focus group outcome. Then six weeks later, a core (turnkey) group of the original panel, again demographically weighted, would be re-interviewed. Then and only then would the report be written up and the metadata considered.
A million miles away from the dire Focus Group lite approach adopted these days. The problem being, the groups are not giving out meaningful data, which can be used, but just a series of random unconnected 'starbursts' as we used to call them. No wonder the election polling is all to pot! It's not likely to improve either...
E&OE google.com/+JulianBray Tel:+44 (0) 1733 345581 .