Data, data everywhere - how much do we use?

There's a ton of data out there about your radio listeners. But how much do you use?

It turns out that Pandora has a "data journalist". She's called Liv, and she writes in AdWeek this month about the data stories she's seen on the platform.

It's interesting noticing the differences. Pandora thrives on the long-tail: music that is "ignored on terrestrial radio" - possibly because terrestrial radio is doing a good job playing the hits. There also appear to be a large amount of Nickelback fans on Pandora, too, which just goes to show that not all Pandora listeners enjoy music. Only joking, Chad. Or am I?

The idea of a 'data journalist" is, at first glance, self-indulgent dot-com rubbish: not helped by some of the more ridiculous titles that these folks call themselves. For a while, last.fm had a "data griot", a man called Matthew, who wrote a long piece about what a "griot" meant.

But it turns out that this kind of work is rather valuable: because the data that radio stations gain every day is really fascinating. It just needs someone to make sense of it all.

Ten years ago, working for a national radio station in the UK, I was wondering whether anyone actually looked at the radio screen while they were listening. We've plenty of ways of getting information there - RDS, DAB's scrolling text, or the radio player on the computer; but did anyone look?

I put a competition on there - not mentioned by anyone on-air, and with different codes per platform. It turns out that people did read the radio screen: but in a different way to how they listened on the radio. They read the screen - and acted on it - in the evening; even though most radio listening happens in the morning. The amount of competition entries at 10.00pm was about the same as at 8.00am - yet the actual audience at 8.00am is much, much higher. It turns out that breakfast is a bad time to expect interaction from your audience.

Useful data, a useful story, and useful to earn revenue from - as we used the evidence to go out and sell off-peak advertising.

The kind of data we get from online streaming is even higher. Log the listening times, location, the audio volume, and when people tune out, and you've tons of fascinating data that can tell you stories. Companies like Voizzup and RadioAnalyzer specialise in using this data for all kinds of programming insights. Monitoring the volume control is especially interesting, since for talk radio it appears to correlate to "I find this bit interesting".

Do, while we might - legitimately - scoff at the title of a "data griot' or a "data journalist"; there's actually quite some benefit of this kind of role at your company.

What could you learn from the data you already have?