Junior Masterchef makes a muted return

As they paraphrase in the classics … No shit (that) Sherlock.

Hi Campbell,

In its favour – it’s every day, which arguably makes TV the most accountable medium.

Against: It’s based on a sample, and of course doesn’t measure what that sample is actually doing in the room at the time.

So it’s probably pretty robust as a buying currency and a means of juding how one show has done against another. But of course it can never be precisely right about how many people actually watched.

Cheers,

Tim – Mumbrella

Hi Campbell,

As Tim points out, the TV ratings are a sample. Mind you a pretty big one (around 13,000 people a day with every minute of the day reported for metro and regional). Where the sample can struggle is with some of the very small channels – e.g. community broadcasting, subscription TV tier-only channels – but they tend to be bought on the uniqueness of their niche target rather than the size of their audience.

To put the sample size into perspective the major pollsters correctly predicted a hung parliament before the recent Federal election based on sample sizes of 1,000-2,000. Yep – we put more sample size into TV than elections!

Also, the sample data is closely interrogated every day – we know that not everyone pushes their button correctly every day so up to 10% of the sample is removed because of data that just doesn’t make sense (like leaving the Foxtel box on the cricket all night after you’ve gone to bed – false tuning).

So, all in all, TV’s measurement is very much world class. Of course, it just enumerates the audience – it tells you nothing about why people watch shows, what they think of them (though watching them episode after episode gives you a pretty good clue), though it does tell you if they turn off part the way through (another good clue).

In contrast to other media TV is ‘way, way up there’. For example, I often hear that online is the most measurable medium. That it is. It is also the least measured “audience” of any medium. It has historically focussed on hits, page impresssions, unique browsers etc. All of those bits and bytes in the data stream is one computer talking to another one. We as users make the assumption that a computer = a person and it very often doesn’t. We also assume that a cookie = a unique person and after a week or two that is very much the exceptionand not the rule. The data stream can make assumptions about the demographics or ‘who’ is in the data stream, but they are more often that ot exactly that – assumptions. That is, online measures the quantum of traffic very well, but not so much the audience. Conversely, TV is very good at measuring the composition of the audience but as the medium continues to fragment it will come under increasing pressure measuring all the fragments of that quantum. Fortunately there are sophisticated statistical data techniques that allow this to be done with a great deal of robustness, which I suspect will start rolling out over the next few years.