Daily Sun a victim of lag, not poor research

When SAARF AMPS 2003A released the first 12-month readership figure for Daily Sun, some concern, rather than jubilation, greeted the paper's amazing growth in readership. How, it was asked, could a lower LSM paper have only seven readers per copy, when other higher LSM papers were racking up as many as 15 readers a copy? Surely Daily Sun should have reported far more readers?

Some commentators wondered whether a mistake had been made in the data. But inaccurate research would have been reflected across all media, clearly not the case if you consider the relative stability of the media consumption patterns shown in AMPS and RAMS 2003A.

Part of the answer lies in the frequency of the AMPS research. When the six-monthly AMPS and RAMS surveys were introduced in 1999, the industry decided that the official currency would nonetheless remain 12-month figures.

This was a good decision, because if six-month data had become the new currency, many media owners would have been constantly reprinting their rate cards, as their figures rose or fell in one survey, only to normalise in the next. The 12-month's rolling data therefore smoothes out non-significant fluctuations and more clearly reflects real trends.

Twelve-month data prevents this seesawing from happening, smoothing out random blips to produce a more stable picture of media consumption patterns over time.

For both AMPS and RAMS, the latest six months of data (January to June 2003) is taken and combined with the previous six months (July to December 2002). The data is then re-weighted to produce the final 12-month rolling data for SAARF AMPS 2003A. This produces a large and stable database which is used as the currency for the buying and selling of media space and time.
There is however, as with everything else in life, a trade-off to the releasing of 12-month and not six-month data - any medium that is growing apace will not have its full growth reflected, because its lower figures from six months prior dilute the current figures. The opposite applies to a medium in rapid decline, whose readers per copy might initially not decline as quickly as the circulation.

This is precisely what is happening with Daily Sun. In the months following its launch, Daily Sun had a readership of 383 000 (July to December 2002). The following six months proved just how incredibly quickly the paper was growing, with a readership for January to June 2003 of an astonishing 1.445-million! Once these figures are combined for 12-month data however, the paper's readership 'drops' to 874 000 over the full 12 month period. Due to this diluting effect, when advertisers buy into the paper now on 12-month's data, they're actually getting much more value for their rand, since they are getting far more readers than they are paying for. Remember too that even the latest six-month data is now already, on average, five months old, and all indications are therefore that the current readership will be even higher!

All media measured on AMPS suffer from this lag effect, but because most publications are relatively stable, the lag is most often not noticeable. Media whose audiences are rapidly changing are most affected although the effect decreases markedly as the new player's circulation stabilises.

Daily Sun is, therefore, not an isolated case. In fact, the same phenomenon is seen with Sunday Sun, as it continues to grow significantly survey on survey since its launch in July last year.

The paper's six-month readership figure for July-December 2002 was 1.371-million, and 1.824-million for January-June 2003. As a 12-month figure however, its readership is the somewhat lower 1.587-million.

A very important fact, however, can be clearly seen from the figures of the Sunday Sun. Sunday Sun's readership growth is in line with its circulation growth, and over the last three surveys, readers per copy have grown from 9.2 to 10 to 11.3.

There is no reason to doubt that Daily Sun will not also follow a similar pattern, and over the next few surveys, will see its readers per copy figure steadily climbing, as readership lag catches up with circulation growth.

Unfortunately, there is no easy way to lessen the lag effect, short of increasing the frequency of the AMPS research, which is prohibitively expensive. In that wonderful world where money is no object, it would easily be possible to publish research results with greater frequency, therefore enabling us to track media on a monthly basis if required. However, for most of our print media this is not essential.

AMPS and RAMS data is nonetheless an accurate picture of reality, as is born out by the consistency of the results of virtually every medium measured on AMPS. Most importantly, these studies provide comparable figures for competing media which makes it possible to compare apples with apples.