Is the money in the meta-data?

Home / Is the money in the meta-data?

My timing is obviously improving.

I ended my last post questioning the risks of broadcasting meta-data over the air, and how it might be used to create websites and activities outside the control of the broadcaster. I really do need to thank the good guys at Absolute Radio for launching their Compare My Radio site last week, because it’s a real example of how this can happen, and how it should be a point of discussion in the industry.

Compare My Radio uses a series of bots to scrape the “playlist” or “just played” pages of various radio station websites, work out the title/artist information, and pipe it into last.fm Then they use last.fm’s investment in statistical analysis to work out which stations play which artists at what frequencies, and merge that all together into their website. You can look at station’s “variety” index, or ask which station plays your favourite artist most often.

A couple of days later, Bauer’s feed for KISS was reported by the site as being broken, and a discussion ensued on Twitter about whether it was deliberate or accidental, and if it was deliberate, whether or not it was a reasonable response to a competitor farming their playing now information in such a way. As it turned out, Bauer had deliberately broken the feed because it was completely failing to represent KISS’s variety correctly, as their specialist shows aren’t played off playout, don’t appear in the website, and therefore don’t make it into the last.fm analysis.

But was their response reasonable?

To answer that, let’s have a look at the business of radio. Radio, as a medium, has lots of listeners – as many as its always had. (A little older maybe, but heck, the whole population is ageing). The problem facing commercial radio is that share of adspend is under real pressure, with more money being diverted to on-line which is perceived as being more accountable, even if its effectiveness compared to radio is open to a lot of discussion.

How do you counter that? I think you do it in two ways:

  • Reduce your reliance on classic airtime revenue
  • Make radio advertising more measurable, accountable and interactive

Meta-data plays a critical role in both these changes.

Reducing your reliance on classic airtime revenue

A fact lost on some media analysts is that “the Internet” is not a medium, it’s just a transport. It’s quite possible for a radio station to counteract declining airtime revenues by ramping up on-line revenues. It’s still a radio business, just a business using broadcast and internet for its content distribution model.

So what draws on-line crowds to your website? Obviously content, but in this search-engine dominated world, and with a burgeoning number of connected appliances, it’s not the content that gets you traffic. It’s the description of the content – the meta-data – that gets you Google juice and rankings in Bing and clicks from passing traffic.

But what if your data is being grabbed by Compare My Radio, and they’re aggregating it with everyone else’s, and getting massive search ranking and authority for Artist and Title searches? That’s your fault, says James, for not building your site right. They’re not selling any ads on their site, so what’s the problem. (To which I answered “yet”).

And what happens if someone starts creating e-commerce opportunities from your station, and others? And again, getting that SEO authority. It’s taking traffic, clicks and e-commerce revenue away from your site.

Doesn’t matter, says Matt. As soon as you put meta-data out there, it’s free (as in libre – public domain). I disagree, and there’d be a huge problem in general if anything that was broadcast immediately became public domain. (Should it be legal for me to download a song from iTunes at 79p and sell it on my own website for 89p?). And the pages that Compare My Radio scrapes definitely have a copyright statement on them.

Making Radio Advertising more accountable, measurable and interactive

You also need meta-data to know what adverts your audience are listening to, responding to and interacting with. There’s potentially a huge amount of value in that data, and losing control of that could be catastrophic. It’s annoying to lose a couple of pence on each track sold in iTunes, but life threatening to lose out on whole campaigns because someone else isn’t passing meta-data to you, or claiming bounties for listener referrals.

Meta-Data Lockdown?

I’m not advocating keeping meta-data under lock and key. It’s pointless (as pointless as trying to stop people having digital music), and hinders lots of fun and creative ideas that could generate lots of interest and value in radio.

But meta-data belongs to the creator, in exactly the same way as the content it describes, and they have to remain part of any value chain. And that means having some control. (Yes, I said it, the “C” word).

It seems reasonable to licence meta-data out to people, and it’s entirely feasible to make that a zero-cost licence. Indeed, if you want, you can have something called a FRNDZ (Fair, Reasonable, Non-Discriminatory and Zero-Cost) licence, which means that anyone who sticks to your Terms / Acceptable Use Policy can have a go. It’s exactly the way Google lets people use Google Maps in their own sites. You tick the box, we give you an API key.

If you’re going to have a licence, you need to make it easier for licensed users to get the data than those who haven’t got a licence (otherwise, why do it?). So (paradoxically?) I’m actually suggesting that radio stations produce higher quality meta-data feeds for their licenced users and conversely, make it as awkward as possible for those who won’t sign a licence to get decent data.

I would be cautious about how much machine-readable information you broadcast without any controls, but provide a route for innovation and experimentation that might just unlock new value for you. That will reduce your reliance on traditional revenue, and bring ears and clicks to your station.

The team at OGS Labs are clever technologists, of that there’s no doubt. But I think, with Compare My Radio, they could have done better if they’d spoken with their colleagues and asked nicely if they could share some data, rather than sneaking up and stealing it away.

Comments

    Paul Webster

    When the Kiss data was removed wasn’t the stated reason that the information was incomplete (because Now Playing info is not presented for all tracks, especially specialist shows) and hence was giving a false representation of their output. An implication being that they would not object if it was complete. Were they being as economical with the truth?

    Reply
    Richard

    A recent courtcase in Australia ruled that future TV programme listings are not protected by copyright.
    ( http://bit.ly/obs0Z )

    It would be interesting to see what the verdict would be in the UK if Radio stations tested whether past track listings are protectable by copyright.

    Reply
    Paul Webster

    The reason stated by Bauer could be read to suggest that once they have fixed their meta-data so that it accurately reflects their content then it would become available for such mash-ups again. Will be interesting to see.

    I had not realised that last.fm was being used as the repository for the data. Were they aware that they are holding potentially contentious data?
    I know that a few stations do submit their own tracklistings there – so maybe OGS could open up their comparison to all such stations – and also allow last.fm users to compare their own profile with stations – then it starts to be a more useful service.

    Reply
    Phill Clark

    I’d agree totally with the concept of ‘Control’. We have no issues as a business with this content being used, however it would be nice to be able to control a clean feed for this data rather that it rely on scraping pages or scripts which weren’t designed to have this done to them.

    Personally, I believe that if it remains a free service with no adverts or ecommerce options then why not, in fact I’d love to see and API from the service soon so more mashup concepts can be created from all the data and analysis.

    Reply

Leave a Comment