In the media, measures of attention are big business. They determine what you can charge your advertisers and how seriously people take your outlet. In commercial TV and radio, stars and shows live and die by their figures.
But the figures are often far from perfect. Of the many official ways audiences and reach are measured in the Australian media, a range of methods and biases can throw things out. Some rankings are freely available to anyone who wants to look them up. Others are kept in-house, or sold to subscribers for a fee. Some you can take at face value, and others, well, it doesn't hurt to be aware of their flaws. Here are the most common measures of audience used by media companies, and what you should make of them.
The small screen
The TV figures are some of the most important of all, partly because of how often they come out and how (relatively) robust they are. On the small screen, how many people are watching is measured minute by minute. TV execs can pore over the previous day's ratings from early the next morning. Things that don't rate well don't tend to last.
The company that wields so much power over the small screen is OzTAM (short for Australian Television Audience Measurement), which is owned by the Seven, Nine and Ten networks. It installs measuring hardware on the televisions of several thousand homes (5250 are participating in the free-to-air measurement system in 2017) and simply measures what that sample are watching, night by night, minute by minute, demographic by demographic. You can get access to them by subscribing (a small number of industry players pay millions a year to do so), but they're not hard for media reporters to get their hands on for free, which is why you see them reported on in so many places (including Crikey).
If the sample is representative (OzTAM weights the survey so it is representative of the broader population), then the measures are statistically robust, if not perfect -- it is, after all, an extrapolation on a smallish sample. But television is being disrupted, and that's starting to show. Many young people no longer own televisions. Australians of all ages watch things online rather than turning on the TV set. Streaming figures exist, but not in a centralised place you can check the next day. Many are tracked internally by the various broadcasters, and not publicly released.
[Playing in traffic: the dirty tricks publishers use to boost online views]
Another problem with streaming is that people stream TV when it suits them, not the moment it's available. The overnight TV ratings are easy to understand and report on, but for many types of shows, they do not capture total viewing. For dramas in particular, the TV ratings are less than robust. Last year, Crikey pointed out that the ABC's Cleverman, a new indigenous sci-fi drama, got killer reviews, but poor ratings. But our analysis didn't include iView views. We did ask about them, but the ABC doesn't routinely release such figures, and it didn't release them to us.
Multi-channelling has also complicated things -- some networks run shows on more than one channel at the same time, which can provide an audience boost. Still, for all their faults, the OzTAM figures are timely, detailed and relatively transparent.
Don't touch that dial
The radio ratings are just as closely watched as those in television. But unlike their cousins in television, radio execs have to wait.
A few times a year, thousands of people across Australia are given diaries and told to fill them out over a given week. Every radio ratings survey averages out responses over a few weeks, and GfK, which since 2013 has complied the ratings for industry body Commercial Radio Australia, carefully controls the age and geographical distribution of those selected.
Minimum sample sizes for each city, from GfK's explanation of how it does the radio ratings.