Here is my perspective on the ratings vs recall debate. At the end of the day, good business decisions derive from accessing many information sources. Media decisions shouldn't be any different, and certainly should not solely derive from a currency based on consumer memory.
But first, a close-to-home example of the quirkiness - and danger - of recall surveys. According to Pax, AXN is a top 10 channel in South Korea, based on past seven days' viewership. This finding delights us.
Here's the caveat: AXN has not yet launched a Korean service, except for US military bases. The 'evidence' doesn't seem to fit the whole 'truth'.
The regional syndicated studies (Pax and ATMS) focus on a small group of elite Asian viewers. The respondents travel a lot, which may partly explain AXN's unexpected performance in Korea. But the regional recall surveys cover the top 10 to 20 per cent of affluent Asians, whereas pay-TV penetration goes much deeper, varying from 30 to 90 per cent of TV households. By skimming the top, recall surveys do not reflect the full performance picture of regional channels. Second, both syndicated studies are based on recall. No detailed viewing data is available. The surveys are more of a brand 'stature' indicator rather than a measure of a channel's reach and viewership. While this brand 'stature' indicator is useful for my business, it's less relevant in media buying situations. There is a known gap between actual media consumption and 'recalled' media consumption.
Finally, Pax and ATMS miss the teen and early 20-something audience which many marketers target.
Now let's look at ratings. Peoplemeter data provides accurate and informative rating and viewership information, including timeblock rating, programme rating, and even minute-by-minute ratings. It captures viewers' detailed TV consumption patterns 'live' rather than relying on memory. Peoplemeter data allows for demographic targeting, with continuous monitoring 24x7x365, and provides a local context for channel performance. The limitation is that only seven Asian countries have pay-TV measurement: Singapore, Taiwan, India, Hong Kong, Philippines, Korea and China.
Secondly, as with all research studies, there are issues of methodology.
The sample sizes are sufficiently large on a country basis, but when narrow demographic targets (say, females 15 to 20) are sliced and diced, sample sizes can get thin. Rex Stout famously said there are two kinds of statistics: the kind you look up and the kind you make up. Media practitioners should concentrate on the evidence, but be wary of just using the evidence which consumers make up.