Blogging Nick Piggott

Apple iPod Nano with FM (C) 2009 Apple

Just when you think there’s nothing interesting you can blog about, Apple come and chuck fresh meat to the wolves.

Of course, everyone’s excited about Apple including radio in one of their devices for the first time. That’s clearly good news. It would be amazing news if it was a DAB Radio in Europe, and an HD Radio in the States, but let’s work on that one. Baby steps.

Let’s assume that Apple don’t incorporate functionality into their devices unless they think users are going to go “wow – cool”. As Mark Ramsay says, Apple didn’t just throw an FM tuner in there; they “enhanced radio”, so it includes pause/rewind and tagging. Adding this kind of functionality costs real money (in material and engineering time), so we should be pleased that Apple see that as a worthwhile investment. Yes, Radio is still cool, and still valued even by the cool kids who buy Apple iPod Nanos. This is a “radio” that 15-24s will love to have.

James explains a bit about how the existing Apple iTunes Tagging works. It’s a system designed to do one very specific job, for one specific group of stations and listeners. It transmits Apple iTunes Catalogue IDs in spare RDS ODA (Data) groups, using a form of encryption (discuss…). The radio station incorporates the iTunes IDs into their FM RDS transmission, the iPod Nano receives/decodes this, and when you hit “Tag” it stores the ID/Artist/Title in memory. When you sync up your Nano with iTunes, iTunes converts that into proper store links, and offers you the downloads. It works. Listeners can tag songs on the radio, and buy them in iTunes. A similar service is also available on HD Radio, and was launched earlier, IIRC.

So what’s not to like. Isn’t this the perfect demonstration of innovative revenue generation in a digital media world?

Maybe, but I don’t think it was initially designed with the listener in mind. It looks like a system designed to turn radio listeners into Apple iTunes customers. There’s nothing wrong with that, incidentally. The rather depressed radio business got a big kick out of being able to announce a tie-up with Apple, who are highly regarded. There’s significant kudos is being allowed to play with the smartest boys on the block.

James has pointed out the weaknesses in the existing system. It doesn’t scale terribly well (although HD appear to be also transmitting different tagging information to support Microsoft’s new Zune HD), and it only works for iTunes and material that’s in iTunes.

There another weakness in the system, in my opinion.

If you look at how the meta-data moves around, it goes in one direction only. From the radio station, via FM, the Nano, iTunes and to Apple. After the radio station has splurged the meta-data out on the broadcast platform, it has no control or visibility of it from that point onwards. There has to be a contractual relationship between Apple and each Radio Station for Apple to pass information about the songs sold back to the radio station. I have no idea how detailed that information is. Does it list every transaction, by every device, by time of day? Does it report transactions, or tagging events, or both? Or do they just get a $ total each month and a cheque for the affiliate fees?

Excluding the broadcaster from the process, and obfuscating the outcome, diminishes the value for radio. It turns us into an customer acquisition vehicle, without getting rich information on listener behaviour.

There’s also the small problem of ne’er do wells “stealing” the meta-data. Let’s assume that someone nefarious decides to strip that meta-data, and amend the affiliate ID to be their own. You might use an apparently legitimate streaming portal, or attractive device, and that money would go to the middle-man, not the radio station. The value of meta-data is increasing, and we should be more careful about whom we exchange it with. In my opinion, broadcasting meta-data risks destroying value. I do agree that meta-data should be open, but I generally think that you should know who you’re providing it to. (I’m going to blog about the side-effects of this shortly).

As you’d expect, I think the RadioTAG model is fairer. It keeps our meta-data relatively secure, whilst still allowing legitimate users (like listeners and Apple) to have access to all the information they need. It scales well, because it’s not transmitting vendor specific information over the air. The broadcaster can see who is requesting what meta-data when, and use that to track listener behaviour in real-time. And very importantly, it lets people tag anything interesting they hear on the radio, not just the songs.

I’m excited that Apple are into radio. I’m excited that the Nano is such a great little device. I’m excited for the prospects of Tagging on the Nano. I just want to make sure we make it great for listeners, as well as for radio stations and for Apple.

대한민국 터치다운 - McCarran Int'l Airport, NV USA (CC) gTarded at Flickr

I’m enjoying Malcolm Gladwell’s latest book, Outliers. In the same vein as Freakonomics, it looks deeper into why certain people or events deviate from the norm, become exceptional – why they are “outliers”. It’s helped create a new view on something that has puzzled and frustrated the world of Digital Radio, and handily includes references to Aviation and Korea, which is where we start.

One of the chapters in the book is called “The Ethnic Theory of Plane Crashes“, and describes the infamous crash of Korean Airlines flight KE801 into Guam in August 1997. (There’s an episode of Air Crash Investigation that covered it – Part I, II, III, IV, V). To summarise, the pilot flew his 747 plane into the ground, despite the instrumentation and the crew being well aware of what was going to happen in good time to be able to avoid the accident. 228 of the 254 passengers died.

It happened at a time when Korean Airlines was putting planes into the ground with worrying frequency. In the book, Gladwell says Korean Airlines was crashing planes 17 times more frequently than United Airlines. This wasn’t due to badly maintained aircraft, or dangerous conditions – the crews just kept crashing their planes. (I don’t know if “17 times” is accurate, but I do know that it’s a well established rule in flying circles never to fly KE metal, and one I have myself stuck to rigidly, preferring to route LHR-FRA-ICN on my trips to Korea).

The obvious conclusion is that Korean Airlines crews were incompetent, but Gladwell suggests it wasn’t incompetence – it was deference. The Korean culture is so deferential to figures of authority or power, the  members of the crew who could see danger increasing simply did not feel that they could bring it to the captain’s attention. It was socially unacceptable for them to question his judgment, or even infer that he wasn’t fully aware and in control of his aircraft. Looking at the transcripts from the flight recorder, it’s excrutiating listening to the First Officer hint and suggest to the Captain that they might actually be flying straight into the ground. Only when there’s less than 7 seconds to impact does the First Officer clearly call for a “Go Around”. Too late.

This degree of deference can be measured – the PDI (Power Distance Index) measures the degree to which people are deferential to figures of authority or power. The PDIs of many nations (cultures) have been measured, and there is a remarkable correlation between  the amount of deference in a in culture, and the plane crashes in those countries. It seems to be that such a degree of deference negates the value of having subordinates to help provide vital input and monitor situations.

So where’s the connection with Digital Radio?

It turns out that there’s another nation with a high PDI. France. France has a  PDI value of 68 – it’s a highly deferential nation. (In context, Britain and Germany have values of 35, and Austria just 11).

All of a sudden, things are clicking for  me. Here’s why.

France chose a non-standard version of DAB – a cut-down version of mobile TV (basically, mobile TV minus the video, or with very little video). Something that the Koreans (them again) invented and dubbed T-DMB. The most prominent figure in that decision for France to the use Korean originated T-DMB system was a man called Sylvain Anichini.

M. Anichini was the Director of Technology for Radio France – in hierarchical terms, he was pretty much at the top of the roost in French Radio. For whatever reasons he had, he became a passionate and vehement supporter of T-DMB. And he would give not a moment to anyone who didn’t agree with his decision. It’s maybe understandable that he repelled approaches from the other DAB nations, on the basis that his “sovereignty” might be undermined. I believe he stormed out of more than one WorldDMB meeting, and was very insulting in public session to a number of fellow professionals. It might also be the case that M. Anichini found dealing with Koreans, and their deferential culture, much more appealing than dealing with those apparently insolent and disrespectful English and Germans.

What’s interesting is the number of people within the French radio industry who privately disagreed with both M. Anichini’s decision, and his behaviour. I’m aware, both directly and indirectly, of heavy sighs when discussing the path being followed, but I (and others) could never understand why those people who were uncomfortable with the direction taken would never raise the issue more openly. I was obviously failing to understand the immense gulf between our culture and the French culture, and what appears to be a overriding and almost smothering cultural barrier to challenge bad decisions made by a person above you. (Maybe this is also why the French are so apparently tolerant of the dalliances of its politicians?).

What was happening in France with DAB had parallels with what happened to flight KE801. One man, whether intentionally or just out of disorientation, was about to bring everything crashing into the ground, and his subordinates and colleagues could not do anything to stop it.

Of course, with Digital Radio, nobody got hurt (apart from a bit of pride) and nobody died. Compromises were made, and positions adopted that allowed the French decision to accommodated, albeit at a financial cost for the entire DAB community worldwide. M. Anichini left Radio France at some point after the decision was taken, but before licences were awarded by the CSA.

So what did Korean Airlines do to turn around their appalling safety record? They hired an American and enforced the use of English in the cockpit, as a way of breaking down the deferential barriers created by the Korean language. It seems unlikely that the same approach would be appropriate for Digital Radio in France.

So here’s my brief contribution to the flurry of analysis of Lord Carter’s Digital Britain report.

The biggest news is that we get a target date for switchoff (sorry, “Digital Upgrade”). 2015 is the year we should be flipping the OFF switch on (almost all) analogue radio, and offering universal coverage of DAB. That date can now be plugged into business plans, and financial projections, and hopefully provide the necessary laxative effect to the recently sluggish developments around DAB in the UK.

So, rather than dissect all of the Radio section of the report, which others will do better than I, here are the bits I particularly noted:

It’s a full switch-off (“upgrade”)

Some summaries have suggested that the 2015 deadline only applies to national radio. It doesn’t – it applies to all services being carried on both national and local multiplexes (3b.10). The only thing left on FM post 2015 will be very small scale services; either commercial or community. There is not going to be a dual-speed changeover, which leaves local radio dragging along for years with a foot on each platform. That’s good.

Support for WorldDMB Profile 1

There it is, snuck away in 3b.20 – receivers sold in the UK should be at least WorldDMB Profile 1 compliant. The box on the following page is a little more explicit in saying that we are giving ourselves a migration path to DAB+, which is the smart thing to do. Nobody seriously considers DMB-A (the Frankenstein bodge invented to make an ill-informed decision seem at least slightly less ridiculous) for radio, so let’s ignore that. Some commentators have, incorrectly, said that Profile 1 includes DRM. It doesn’t, and DRM needs to mature a great deal more before it can earn a guaranteed place alongside DAB and DAB+.

Improving Signal Quality

It’s no secret that I don’t believe DAB should be crippled by being forced into universally super-serving a small fragment of the audience that expects ultra-high-quality audio from every radio station. The market can and will decide what audio quality is right for which stations and bearers.

But I do believe that we need to offer robust indoor and handheld coverage to everyone who currently enjoys that from FM now, and by crikey, it’s not rocket science to do it. Australia’s got the right idea – power. And more of it.

There’s some more crypticness in the report. It talks a lot about achieving equivalent coverage prior to 2015, but only in 3b.23 does it explicitly recognise that indoor coverage must be more effective. It also recognises that there’s some cost in achieving network upgrades, but notes that there is opportunity for negotiation between the BBC, multiplex operators and transmission providers. That’s timely, as many of the initial multiplex transmission contracts come up for renewal soon, and knowing with certainty that it’s worth spending money on the infrastructure is very valuable.

Replanning the network

This wasn’t as explict as I had hoped for. There is reference in 3b.26 to giving OFCOM the powers to re-plan and amlgamate multiplex areas, but I would really would like to have seen a more definite commitment to re-plan at a spectrum level to get a step-change in coverage (up) and costs (down). At least there’s a statement that sorting out coverage shouldn’t be as expensive as some people might have made out it could be.

And now – drum roll – the best bit…

In fact, it’s so good, it’s the only bit I’m going to quote verbatim from 3b.31:

Functionality and interactivity must become central to the DAB experience.
EPGs, slideshows, downloading music, as well as pause and rewinding live radio
must be developed and brought to market on a large scale. Broadcasters and
manufacturers must seek to develop and implement digitally delivered in-car
content, such as traffic and travel information.

Well, we waited a decade, and now it’s a formal part of the plan to digitisation. Digital Radio must prove its worth by doing something… digital. If we don’t use the platform and spectrum we’ve been given (and will continue to get for free for a while – 3b. 27) to evolve radio, what’s the point of doing it? Same value, different platform?

If the other parts of Digital Britain are designed to create confidence in building transmission infrastructure, and writing long-term financial plans that support transitionary investment to achieve that, then this is the statement that should create the confidence in investing in a new kind of digital radio, and it’s about a content led experience that’s enabled by a universal, free-to-air technology. If the rest of the report stabilises the ship, and gives it a shove in the right direction, this is the bit that signals the start of true innovation and digital change for radio.

95.8 Capital FM iPhone Application

My joining GWR Group coincided with an explosion in the use of music research to decide what songs got played, and how often. The data drove a new format – the “Better Music Mix” that rolled across Southern England in the mid and late 90′s. Hundreds of listeners were surveyed every week to track their changing interests on a track-by-track basis.

Despite that intensive process, there was one thing research couldn’t do. It couldn’t tell you if a new song was going to be a hit with the audience or not. Only after people were familiar with a song could they give you an opinion – virtually all new songs scored badly, simply because they were unfamiliar. The only way to see if a song was popular or not was to take an informed decision, use a bit of “gut feel” and start playing it – albeit gently at first. After about 6 weeks of exposure (assuming a few other stations were also playing it), and you’d start to see the opinions form and polarise, and you could decide to bin it or stick with it.

That experience from the analogue world is equally applicable digitally.

A lot of Digital Radio’s attributes are simply extensions of analogue radio; more stations, improved sound quality, better reception, easier to tune. They all address familiar radio functionality that listeners have found wanting in analogue. It’s not surprising, then, that these are the messages that have most impact with listeners when they’re thinking about reasons to go digital. And in turn, these become the headline messages of a marketing campaign for digital radio.

It’s interesting to compare the motivators people have for purchasing a digital radio with the attributes they say they most value having bought one. Pre-purchase, the concept of text information scores virtually nowhere – nobody buys a Digital Radio to get text information, and it seems to be an utterly valueless attribute. However, post-purchase, it soars to be one of the top five things that people love about their Digital Radios. Before they experience it, they can’t understand it, and so can’t value it. It only takes a short experience to get the benefit, and, even more interestingly, for it to become a differentiating factor between radio stations. Shortly after the launch of Core, research showed that listeners loved the real-time text information on the display, and absolutely slated Radio 1 for not doing the same. (Annoyingly, the BBC fixed that far faster than we expected them to).

We were lucky that text was a de-facto inclusion on almost all digital radio devices, even if the implementation is pretty ropey, on poor displays. (If anyone can show me a DAB Digital Radio that implements the “Clear Message” command in DLS, I’ll be amazed).

The problem is that we need to go further in using Digital Radio to create new functionality and better differentiation between analogue and digital, and between digital and on-line streaming services. And a further problem is that our audience won’t understand what the heck we’re on about until we show them.

I did a demo to the GWR Board in ’98/’99 (along with Dirk Anthony) of our concept for a classic rock radio station called C-Rock (yes, ha ha). The demo consisted of an audio CD, brilliantly imaged by Scott Muller, and a series of HTML 3.0 webpages, which advanced using HTTP META REFRESH tags (this was the 90′s – AJAX was still a bathroom cleaner). It demo’ed our vision of what Digital Radio should be like – a fusion of audio and images. Of course the audio bit of that demo became Planet Rock, and very successful it is too.

But the visual bit of that got stuck for a decade. Listeners couldn’t understand it, so manufacturers wouldn’t build colour screen radios, so multiplex operators wouldn’t allocate capacity for it, and sales teams wouldn’t even consider selling it. Log jam.

Then the Apple iPhone changed that completely.

Quite unexpectedly, the iPhone has provided the catalyst to get visual radio taken seriously. It could (should) have been Nokia Visual Radio, 5 years earlier, but NVR was so horribly badly implemented, it never got any traction. (There’s a moral in there for Nokia, I’m sure). But it’s been the iPhone, and its colour screen that have provided a trial environment for visualised radio, and the feedback from listeners is overwhelmingly positive.

The Global Radio iPhone Apps aren’t the only radio apps that support visuals (although clearly, they’re the best). There’s Absolute Radio’s iAmp, TuneKast and the now last.fm has announced that they’re visualising their player as well. Collectively they’re providing data on listener appreciation (high) and the volumes of visuals delivered, which in turn sizes the commercial opportunity.

I find it ironic that Apple, having kept radio out of the iPhone, has inadvertently provided our best research source yet into a truly innovative change to radio, and one that our listeners could not possibly have understood or valued without experiencing it. Now the initiative lies with the radio industry to implement it and promote it before that innovation gets stolen by the on-line streamers.

A number of articles and blogs have drawn attention to the ability of Google searches to provide early indications of change. Google announced that they were providing information on people searching for infomation about ‘flu to map outbreaks, and this week there was an article in The Economist about how eerily accurately the decline in people searching about Ford cars was reflected in actual sales decline.

So what does Google’s clairvoyance tell us about DAB Digital Radio?

Let’s kick off with the basic trend of “dab radio” anywhere the world.

Google Trends for DAB Radio Worldwide (Click to enlarge)

As a piece of calibration, this seems about right. Not surprisingly, the two countries that have really “got” DAB, the UK and Denmark, are pulling all the hits. And there’s a surge interest around Christmas which absolutely matches what happens to sales. (And Bristol is high source of traffic – can’t imagine why (OK – probably because Virgin Media have a connection to the Internet here…)).

The trend is pretty static, globally – but you can see the growing noise in the press about DAB, which continues fairly unabaited. (No, I can’t explain why Danish is inexplicably the top ranked language. Maybe the pro-rata’ed access to Danish language articles is much higher than to English language articles?).

So, let’s narrow it down to the UK.

Google Trends for DAB Radio in the UK (click to enlarge)

Restricting the analysis to just the UK really don’t change thing very much at all, which probably gives us an insight into how much the volume of queries worldwide is driven by and influence by the UK. I think this means we probably drive virtually all the Google queries for DAB Radio. (More on that in second).  If I remember correctly, 2004 was the first Christmas that the BBC really pushed DAB, probably because they actually had some new radio stations to talk about. My intepretation of the declining peaks at each Christmas is that people need to know less about DAB and need to less searching to find out who sells it. And there is a drift downward in the number volume of queries. Does that mean that people want to know less about it, because they already know enough? Is that too optimistic?

But we know the UK is DAB-happy. What about the other big European country which was apparently so enthusiastic about implementing DAB. How does it look in Germany?

Google Trends for DAB Radio in Germany (click to enlarge)

This looks rather weird. It suggests, from the shape of the graph, that overall query volumes are tiny. I compared the width of the “Country” bar graph (in the Worldwide chart) for the UK (98 pixels) with that for Germany (6 pixels). I know that’s horribly inaccurate, but it indicates that there’s probably about 15-20 times more queries for DAB coming from the UK than Germany. That Bayern comes top of the list doesn’t surprise – but it’s hard to tell if it’s because it’s the Land that is most active in DAB, or just the largest  of the Länder.

As the media seems to be keen to promote Internet versus DAB as the battle of all time, let’s have a look at the relative performance of those terms in Google Trends. Firstly, across the world.

Google Trends of DAB Radio and Internet Radio, Worldwide

Not entirely surprisingly, globally Internet Radio is searched for a fair bit more than DAB Radio. The average ratio is 10.8 : 1, but that seems to suggest that DAB is actually out performing Internet Radio in terms of interest and search terms. Let’s assume that most of the DAB searches are coming from UK, Denmark and Germany with  a combined pop’n of 147m, against a global population of 6.77bn. That’s a much higher proportion of searching for DAB Radio than Internet Radio. (Although people might also be searching for other terms).

The decline in the search volumes for Internet Radio is confusing, given that it’s apparently in its ascendancy. It’s much more apparent than the slight decline in DAB searching we saw in the UK. The only explanation I can suggest is that as Google gets used more by “normal” people, they are slightly less inclined to search out Internet Radio than the more geeky early adopters? Or has everyone got an Internet Radio now?

You can see from the bottom of this graph the country-by-country breakdown, indexed against DAB. (If you index it against Internet Radio, the country lineup becomes Mexico (!), Germany, Netherlands, Brazil, Peru, United States, Switzerland, Canada, Spain, Austria). Germany is interesting – more of that in a second. And you can see that in the UK, Internet Radio and DAB radio are about the same.

So let’s look at the UK in detail – DAB Radio versus Internet Radio.

Google Trends for DAB Radio & Internet Radio in the UK (click to enlarge)

In the UK, the two search times are neck and neck, with DAB just edging out Internet Radio on the basis of the seasonal interest around Christmas. It’s very interesting that the media perception is that DAB is in a ditch and Internet Radio is it – but that’s not what Google’s users are telling us. Notably, the amount of coverage of Internet Radio (the lower graph) is much much higher than DAB Radio, but it just doesn’t seem to be reflecting or driving interest. That does kind of figure – lots of Media noise about Internet Radio, but real people are looking at both.

Finally, a quick trip back to Germany to see how Internet Radio is doing there…

Google Trends for DAB Radio and Internet Radio in Germany (click to enlarge)

No DAB huh? I guess people will look for their radio choice va the Internet then. But still that dramatic decline in relative search volumes for Internet Radio recently. I’ll be intruiged to see what this graph looks like once the Germans have started promoting DAB+ to their population.

So, what can we conclude fro this graph-fest?

• In the countries that have promoted DAB, it seems to be in rude health, and with no significant decline in interest, despite generally negative media coverage in the last year or so.
• Internet Radio doesn’t seem to be growing interest relative to the growing amount of (largely positive) media coverage of it.
• Relative interest in both DAB and Internet radio is declining as more “normal” people start using Google to look for stuff that interests them. But interest in Internet Radio is declining faster than interest in DAB Radio.
• In Germany, people are interested in Internet Radio (presumably to seek out choice) and would probably just as interested in DAB Radio if it were promoted with confidence.

I’m going to keep an eye on “The Trends” and will maybe update in 6-12 months time. (I’ll also hopefully have some first data for Australia, in which DAB search terms rate 0 across the board).

P.S. Just to reassure you that the terms DAB Radio and Internet Radio are what German speakers would search for (well, as much as any British person) I speak enough German and know enough German speakers to be reasonably confident that the results aren’t skewed by the language difference.