thelondonpaper readership issue

In response to the latest UK National Readership Survey (see previous post), Ian Clark, Managing Director of thelondonpaper criticized the NRS. The research is not doing justice to free dailies according to Clark, because of the survey method used and the low number of readers interviewed. Clark said to Brand Republic:

The NRS is misleading and increasingly irrelevant for the free papers when sample sizes remain ridiculously low. There were 312 respondents for thelondonpaper for the last six-month survey, during which time more than 30 million papers were distributed.

The arguments sounded a bit lame, mostly because thelondonpaper had less readers than rival London Lite, although it distributes 25% more copies. London Lite’s Steve Auckland said it was “typical of News International (…) if the sample does not fit, they just want to change the currency.”

I think, however, that Clark might have a point here. NRS is done by interviewing people face-to-face and finding them by calling at their home (see NRS survey method). With some people – older people, housewives – there is a greater change of finding them at home. To make sure the sample looks like the general audience, the sample is afterwards ‘weighted’: some categories count extra. This is an accepted method but is still problematic with some ‘hard-to-find’ groups.

It makes sense that a paper with a very young audience could run into trouble this way. To check whether this is the case, the readership of the major UK papers is compared.

(This is raw July 2007 – June 2008 data. The latest data are not yet available while detailed data on age groups is not available for non-subscribers. I asked the NRS some months ago if I could have data for research, they didn’t even bother to answer.)

Thelondonpaper indeed has the youngest readers, although London Lite is close. Using another method like online survey – or better: combining different methods – would probably lead to different results.

The bottom line is: if you change the method, data will change as well. At the Readership Symposium in Vienna in 2007 a Canadian research showed for instance that readership data for dailies was higher in the online panel than in the telephone panel – for free dailies it was even much higher.

Adding online survey to the NRS would change UK data as well, more free newspaper readers would respond. Perhaps thelondonpaper would see readership increase – but as London Lite has a rather similar readership (as far as I can see) – their readership would change in the same direction.

My hunch is that distribution could explain the difference in readership better. I am not sure how it is now, but London Lite started distribution earlier than thelondonpaper when they launched. This could have an effect as well.

2 Responses to “thelondonpaper readership issue”

  1. Newspaper Innovation » Blog Archive » Australian readership… again Says:

    [...] discussion looks like what happened in London recently. Also in UK the sample and method have been [...]

  2. Stuart McDonald - News International Says:

    I’m glad that this is now being picked up by industry commentators – we have objected about NRS methodology since our first inclusion on the NRS, as you correctly point out it is not conducive to reaching this young, affluent market through in-home interviewing (especially in London which has its own response issues).
    We look forward, as an industry to hearing how NRS feels they can resolve this, maybe using (as you suggested) a mixed methodology approach.