Audience measurement advances on show at WAM Conference

Jointly organised by the Advertising Research Foundation (ARF) of the USA, and the European Society of Opinion and Marketing Research (ESOMAR), the workshops focused on television, radio, print and Internet audience measurement. Some 470 delegates from 41 countries attended.

SAARF reports back on some of the key findings and ideas presented at the workshops.

Television

1. The keynote address highlighted the growing importance of a global marketing approach, and the building of a global relationship between brands and consumers. To fully utilise digital global media via satellite television and the Internet, marketers will have to change their strategies.

2. A number of papers pointed out that the higher the level of viewer involvement in programmes, the higher the possible effect of advertising in those programmes.

3. The Americans are working on the concept of using satellite decoders to identify the channel to which they are tuned, and to gross-up people meter data to these levels.

4. Professor Matthias Steinman's Telecontrol Group has been measuring satellite television in Germany for about a year. The capabilities of this system indicate that they are clearly ready for the exponential growth of the technological era we are entering - the system can measure as many as six television sets and two VCR/DVDs per set. They also monitor children from as young as three years of age.

5. The problem of declining response rates and the pressure that media fragmentation places on sample sizes, was again evident as a universal phenomenon.

6. A paper from Pakistan showed that problems such as telephone suspensions, the lack of mains-electricity, and voltage fluctuations were not unique to South Africa.

7. Latin America is busy with an approach, comparable to PAMRO, to harmonise television data in Southern America across country boundaries.

8. The Netherlands presented a paper on the measurement of television audience appreciation, which measures how much the viewer liked a programme, normally on a scale of one to 10. Since about two in every three Dutch homes have access to the Internet, an Internet panel of 8 000 is used to measure this.

9. Peter Menneer presented a paper on users' satisfaction with people meter systems in 22 countries, including South Africa. He showed that in countries where the service is provided by a joint industry committee, such as SAARF, users were most satisfied.

10. After more than five years, picture matching is still alive and well in Canada, and most of the problems linked to this method of measuring TV audience have been overcome.

11. Lebanon reported on the use of a motion detector to ascribe viewing when movement is registered and nobody is logged in. SAARF regards this as a dangerous approach - if the family dog moved, for instance, it would be measured as a viewer!

12. Paul Donato of Nielsen Media Research, USA, found that a great deal of future viewing in specific households could be predicted by looking at the household's historic viewing behaviour. Donato will be one of the speakers at the SAARF One-Day Media Research Symposium on October 3, 2002, in Johannesburg.

13. Arbitron reported on the progress it has had with the Arbitron Personal People Meter (PPM), which automatically measures television and radio audiences. Jay Guyther will present an update at the SAARF One Day Media Audience Research Symposium.

Internet

Unlike television audience measurement, there have been few developments in online research. Researchers are still investigating the two approaches that were reported on at last year's ARF/ESOMAR Online Media Conference in Athens - site centric and person centric research.

The only new development was that the Advertising Research Foundation (ARF) in the USA has appointed a committee to further investigate Internet research.

Print

1. A paper from Sweden indicated that this country is using frequency of reading as the industry currency. Most other countries use recency rather than frequency, and average rather than specific issue readership. Sweden however, is confident that this method brings users "closer to the truth".

2. A paper from the US however, claimed that the use of frequency was "flawed", as it overstated readership at the higher, as well as the lower ends of the scale. Pass-along readership was also understated by using the frequency question.

3. Another paper from Sweden reported on the trial use of SMS messages on cell phones to measure readership. The Swedes concluded that the method showed potential for future use on a large scale, but that further experimentation was required.

4. In the USA, a pilot sample of TV homes was equipped with a device to link the TV to the Internet - results of Internet measurement of readership looked promising. The Internet penetration in the States however, is much higher than in South Africa, and so this is not yet an option for us.

5. The Germans presented a comprehensive paper on a through-the-book approach to determine the best page, and position on the page, for advertising a broad variety of products.

Radio

1. The keynote address to open the last day of the WAM Conference that dealt with radio, was delivered by Sheila Byfield of the UK. Her topic was "Media under threat? Who will survive?" It was interesting that Byfield felt that print media was not under threat. Byfield has been invited to present this paper as the Wally Langschmidt memorial presentation, which is the Keynote Address of the SAARF One Day Media Audience Research Symposium in October.

2. The highlight of the day was the report on new developments with the Arbitron Personal People Meter (PPM), and the Swiss Radio Control Watch. Arbitron presented results of the first three months' of its experiment in Philadelphia. Belgium, Canada and the UK are also testing the PPM. Radio Control reported on a year's experience in Germany using the Radio Control Watch to measure radio and television audiences.

It was clear from both presentations that these technologically advanced methodologies would introduce a new currency with different audience levels. Both recorded significantly higher audience levels for radio as well as television, compared to levels achieved using diaries. No comparisons with peoplemeter results were available.