A Sticker-Diary Cautionary Tale (Or, As The British Say, "A Sticker Wicket")
Written Feb. 6, 2009 by Larry Rosin in Marketing + Terrestrial Radio with 4 Comments
With the Neilsen/Cumulus Sticker Diary soon to make its debut stateside, it may be of interest to American radio broadcasters to learn of a recent related issuefrom the UK.
While Neilsen does not do the radio ratings in the UK -- the entity that does (RAJAR) also uses an aided-recall sticker diary system.
As many people know, in the late summer of last year, well-known national Rock station Virgin Radio was bought and as part of the deal the name had to be changed. It became "Absolute Radio."
When the Rajar report for the fall came out a few days ago, Absolute had numbers well off from those of Virgin. Did people turn away because they were loyal to Richard Branson's famous brand name? Well, perhaps, but far more likely is that the aided-recall sticker system could not handle such a change.
The method there is to have respondents flick through 62 (!) index cards and find the names of the stations they listen to (and then the stickers are placed into the diary). Absolute was saddled with a card that said "Absolute Radio (formerly Virgin Radio)".
Now imagine flicking through so many cards. Do you think your eye would travel over to the "formerly" part? Or would you just flick past the unfamiliar "Absolute" before you would even get there?
Well, it turns out we can prove the situation -- the ratings period started two weeks before the name change! And, guess what, the numbers dove dramatically while the station was still called Virgin. The problem lied not with themselves but with the cards.
All other data produced for Absolute implies they have at least as many listeners, or more likely MORE listeners than before the change. They are serving more streams, their database has grown, other research shows increases; the ONLY downer is from Rajar. And while aided recall might sound like a boon to a station's chances, in this case with unaided recall, responents would likely have said "Virgin" and Absolute could have gotten their credit.
So our two lessons: 1) Don't believe any reports that Absolute's listening is down, only that Rajar's measurement shows less listening; 2) If you're in a Cumulus/Neilsen sticker diary market, be certain to finish off any name changes/format changes WAY before your once-a-year diary drop.
Reader Comments
Your 2¢, in chronological order — add your comment below.
It is true that the RAJAR radio audience measurement system here in the UK is far from perfect but, as you have explained, one of the biggest challenges it faces is the increasing number of stations available in local markets and the resulting fracture of audiences. We have three layers of radio service – national, regional and local – and the offerings are different as delivered by analogue broadcast (FM/AM), digital broadcast (DAB), digital terrestrial TV (Freeview), satellite pay TV (Sky), satellite free TV (Freesat), and the internet. Additionally, consumers have access to a huge choice of on-demand programming and podcasts, particularly from the BBC. This is a challenging landscape for any audience measurement system.
RAJAR data is collected and published quarterly. For some stations, the published quarterly data is based solely upon that quarter’s listening. For others, it is based on listening across the last two quarters, and for some (very small stations), listening across the last four quarters.
As a national commercial radio station, Virgin Radio had a large enough sample of RAJAR respondents in each quarter for its results to be reported on the basis of one quarter’s listening. However, the results from quarter-to-quarter oscillated significantly, so Virgin Radio’s then owner complained to RAJAR in 2004 (see http://www.guardian.co.uk/media/2004/sep/09/rajarradioaudiencefigures.radio ).
As a result of this complaint, from Q3 2004, Virgin Radio’s RAJAR results have been published quarterly, but on the basis of the two previous quarters’ data. In other words, the quarterly results are based on a two-quarter moving average, reflecting six months of respondents filling in diaries.
The latest RAJAR results, ostensibly for Q4 2008, in fact reflect data collected in Q3 2008 (when the station was named Virgin Radio) and Q4 2008 (when the station was named Absolute Radio). This is confirmed by the “H” entry (meaning data collected half-yearly) against Absolute Radio’s latest results on the RAJAR web site (see http://www.rajar.co.uk/ ).
Therefore, whatever change we have seen in the latest published RAJAR results of Absolute Radio’s first quarter since launch is actually an average of its first quarter and the final quarter of Virgin Radio’s existence. It will not be until the next quarter’s RAJAR results that we will be able to evaluate accurately the performance of Absolute Radio, as the result then will be based upon data collected in Q4 2008 and Q1 2009 (both quarters after the station’s name was changed).
I felt it important to clarify this so as not leave your readers with only half the story. RAJAR methodology is regularly blamed for individual UK radio station results but, in this case, the methodology used (two quarters’ data averaged for one quarter’s results) was specifically requested by Virgin Radio’s former owner. Thus, your statement that “the [RAJAR] ratings period started two weeks before the name change” is not accurate in Absolute Radio’s case.
FYI, Absolute Radio is the only one of the UK’s eight national analogue radio stations reporting quarterly in RAJAR on the basis of two quarters’ data. By market share, Absolute ranks seventh out of these eight stations, just ahead of the BBC’s classical music network, Radio Three.
Grant Goddard
Radio Analyst, London, UK
http://grantgoddardradioblog.blogspot.com/
Great post.
Except, we say "a sticky wicket", which rather ruins your headline.
Still. Jolly good show, eh, what?
James
Grant,
What you say about Absolute Radio's (and before that, Virgin Radios') reporting period is accurate. It is over a six month period, and the change was made for the reasons you pointed out.
However, the reference to audiences falling *before* the name change is in regard to the unpublished and unweighted weekly data that we've had made available to us.
This data shows that prior to any format changes made at rebrand, the audience fell dramatically as soon as new labels were published. RAJAR is only able to make changes to labels in the diaries at the start of a period, so unless you time your rebrand perfectly, this will always be an issue for stations.
We remain strong supporters of RAJAR, but it's important to realise how tricky re-brands are.
Adam Bowie
Head of Strategy & Planning
Absolute Radio
Maybe I don't understand how these sticker diaries work as I have never seen one, but can't the book have TWO stickers for the one station? Have one with Virgin and one with Absolute, whatever the station is know as at the time, people pick the correct sticker. Both count to the same station. Loyal listeners that forget the name has changed still reach for the Virgin sticker. And it counts.
No need for "FKA" double billings or perfectly timed name changes. (just know you are going to make the change during the next period)