Next AMPS will reflect market realities

14 Dec 2009

Both the South African Advertising Research Foundation (SAARF) and AMPS contractor Nielsen predict that any readership changes, which may come through in the next AMPS release - 2009B - will be chiefly due to market realities and not due to the change of methodology from CAPI to DS-CAPI.

With CAPI (Computer Assisted Personal Interviewing), only the interviewer had a laptop. With DS-CAPI (Double Screen Computer Assisted Personal Interviewing), respondents look at a second computer screen during the interview, where all show-card prompt material is displayed electronically. For print, similar titles are carefully grouped together to remove as much title-confusion as possible.

New six-month release for comparison

The latest release in September 2009 (AMPS 2009A) was a mixed methodology survey, using CAPI for one half of the Large Urban component and DS-CAPI for one half of the Large Urban and for the entire Small Urban/Rural component.

The introduction of this improved, more accurate methodology naturally brought about changes to the print currency. Print owners were concerned that once the survey went over completely to DS-CAPI in 2009B, they would experience more changes and yet another currency benchmark.

Research shows that this is unlikely, as there are minimal changes between the mixed methodology 2009A survey and the pure DS-CAPI six-month data.

In a presentation to members of the Print Council, SAARF and Nielsen compared the following databases to examine the impact of DS-CAPI on the six-month filter question and inspect average issue readerships (AIRs) in the six-month database:

2008B (12 months), which used CAPI exclusively;
2009A (12 months), mixed CAPI/DS-CAPI; and
2009A (six-month data), which used DS-CAPI exclusively.

The release of the six-month database was a once-off occurrence and was only for comparison purposes.

Major trends

Full DS-CAPI on newspapers

At a six-month readership level, many titles showed increases. At the six-month filter question, readers first claim which titles they have read, and it is here that DS-CAPI has had its greatest influence.

By aiding more accurate memory recall and reducing respondent confusion, the improved methodology has raised the number of people claiming to read a newspaper, both in the daily and in the weekly segments.

Total six-monthly newspaper readership went up significantly from 77.6% to 78.2%.

For dailies, 52% of titles were significantly up, while 48% remained unaffected by the introduction of full DS-CAPI.

Weekly papers showed more change at this filter. A third was significantly up, 27% were significantly down, and 40% remained unaffected by the methodology change.

“This was to be expected especially for the weekend editions of dailies because DS-CAPI has removed a lot of over-claiming for this sector,” says Dr Paul Haupt, CEO of SAARF.

For average issue readership, DS-CAPI has had little impact, since the full effects of the new methodology occur at an earlier stage in the survey.

In the daily newspaper sector, 71% of titles showed stable readerships, with no change between their CAPI and full DS-CAPI results. Just less than 20% showed declines (only four of the 21 dailies), while 10% grew their readerships.
For weeklies, 53% showed stable average issue readership levels, with no six-month database change over their CAPI levels. Twenty percent showed increases, while 27% showed declines -just eight out of 30 weekly papers
For those newspapers which showed a decline in AIR in 2009A (12 months) over AMPS® 2008B, there has been no further decline with the introduction of full DS-CAPI.

“Market forces, and not DS-CAPI, are at work here,” says Haupt. “We should not underestimate the effect of the hundreds of thousands of people who have lost their jobs over the past year. When people lose their jobs, it has a direct impact on their media consumption.

“Those people who were still reading a newspaper on their way to work or at work are now suddenly sitting at home and either not buying a newspaper or only reading it at a later stage. At the same time, TV viewing has gone up enormously and is competing for readers' time, thus affecting the consumption of other media.”

Full DS-CAPI on magazines
As expected, DS-CAPI had a significant impact on magazine readership at the six-month filter question, pushing up levels to reveal a more accurate picture of magazine readership in South Africa.

Since title confusion amongst magazines always played a role with the previous methodology, Haupt commented that it was gratifying to see the correction taking place in this market sector.

This increase was seen in the mixed methodology 12-month database and continues into the full six-month database. Results from both databases are significantly up on 2008B.

Six-month readership of any magazine went from 53.1% in 2008B, to 62.4% in 2009A (12 months) with a further significant increase to 66.0% in the six-month database.
This increase is based on the growth of 50 magazine titles, accounting for 42% of the (non-subscriber) magazine market. Only three magazine titles declined in the full DS-CAPI release over 2008B, while 55% remain stable.
Out of 14 subscriber magazines, five grew, three declined, and six remained unchanged (full DS-CAPI over full CAPI).

“Club magazines are similar to the weekend editions of the daily newspapers - there was a lot of confusion when they were not all grouped together on the show cards. The DS-CAPI method is much fairer to titles in this sector,” says Haupt.

Average issue readership levels showed relative stability, with 61% of non-subscriber magazines (72 titles) showing unchanged AIR levels. Of the 118 titles, 41 showed increases over 2008B (35%), while only five showed declines.
For subscriber magazines, 36% had a stable six-month database AIR over 2008B, 28% showed declines, and 36% showed increased readership.

Overall conclusions

“With both newspapers and magazines, we need to take the recession into account,” continues Haupt. “At the time of fieldwork, the full force of the economic downturn was upon us. In a recessionary environment, many of the declines seen in print were to be expected. We do not believe that the print currency will experience the same dramatic changes in the next survey, unless market forces intervene. Although some effect will still be seen in the next release in March 2010, the impact caused by the change to DS-CAPI is largely over. The industry can now move on from here with confidence that the new methodology is indeed better.”

Download presentations

The full presentation showing the comparison of the three AMPS periods in question, for each print title, is available for download at www.saarf.co.za, by clicking on “AMPS” and then “PowerPoint Presentations.”