podcast

Podcasting—the platform battle

Picture of young man wearing headphones by Gauthier Delecroix

UPDATE 2022. I got this completely wrong – at least the bit where I decided who will win the podcast battle.

If podcasting is going to become a real business it’s going to have to leave the commercial dark ages behind, evolve some more sophisticated audio platforms and —let’s face it— pick a winner. This is the second of two posts about the evolution of podcasting. Part one is about the explosion of new formats.

How did we get here?

One of the geeky pleasures of the audio boom is the secret knowledge that the whole teetering, upside-down pyramid of podcasting stands on the back of a simple technology that’s nearly twenty years old.

It’s called RSS (which stands for Really Simple Syndication) and it was designed as the simplest possible way to share content on the web —a way to publish lists of stories so that other web sites could receive them automatically. We use the word ‘feed’ when we talk about RSS and that makes it all sound very active — as if you’re firing your content out into the cloud when you create one — but an RSS feed is really just a regularly-updated list with its own URL — a text file that’s so simple you could create one yourself, right now, if you felt like it.

RSS is simple and it’s not perfect — in its origin it was almost the definition of ‘quick and dirty’— but it’s been exceptionally effective. The fabric of the mid-period web – the explosion of sites that became the blogosphere – was essentially a product of RSS. It’s impossible to calculate the number of articles and episodes shared via RSS since the late 90s, and, although you may not know it, you’ve certainly used it — you’ll use it today, in fact.

And the lovely, unforeseeable breakthrough that made podcasting happen was a minor tweak to RSS (thank you, Dave Winer) that allowed publishers to add audio (and video) files to their article feeds. Suddenly you could syndicate your store of audio, turning it into a globally available broadcast channel with hardly any effort. Then, in 2005 when Apple allowed users to paste these RSS feeds into iTunes, the idea of ‘subscribing’ to a podcast was born and the whole thing took off.

But the simplicity that produced the extraordinarily dynamic DIY content industry we call podcasting has also radically limited it. This passive, stateless ‘publish and subscribe’ model means that publishers can know essentially nothing about who is on the other side of the transaction, downloading the audio. So, while creativity in podcast formats and content is essentially out of control, innovation in distribution, ad-tech and in understanding audiences has hardly begun.

The upside, of course, is that Podcastland, at least so far, is a near-Utopia for privacy and freedom fans. The simplicity of RSS effectively inoculated podcasting against intrusive ads, data harvesting, fingerprinting and the rest of the miserable ad-tech toolbox. At least for the time being, subscribing to a podcast is the least compromising thing you can do online. In the increasingly predatory 21st Century ad-tech landscape, RSS is a glorious 20th Century anachronism. Long may it last, cry the listeners.

For podcasters and advertisers, though, the technology is irritatingly non-dystopian. It lacks some really basic features that they’re going to need if it’s going to become a real business, a self-funding medium in its own right.

What’s wrong with podcasting now?

User data is laughably thin. There’s some subscriber data locked up in Apple’s servers but, even if you could get it, it’d be close to useless, at least from a voracious ad-tech perspective. Podcast owners can now get nice-looking listening graphs from Apple and your own server data will tell you how often each episode is downloaded but that’s essentially it. There is no equivalent for the insane level of detail you get from Facebook Insights or Google Analytics.

You’ll learn nothing useful about how people listen. Want to know where your subscribers listen? What time of day? What they’re doing on their phones while they listen? Are they running? Commuting? On holiday? And what else are they listening to? Sorry. Not happening (you can ask them, though).

You can’t serve ads with any kind of intelligence —no pre-rolls or mid-rolls, no dynamically served audio at all— so no holiday ads during the commute, no Bisto ads on the way home and no control over frequency or day part. No geotargeting, no retargeting, no tracking, no spookily accurate robo-ads, no personalisation of any kind. For the time being it’s mattress ads all the way down.

But, of course, this low-tech Eden—this splendid anachronism—is not long for this world. Be sure, the ad-tech is coming, and it’s not just commercial podcasters who want to personalise audio—public service media will want access to these tools too.

The new audio platforms are already here

There’s a generation of new podcast hosting platforms. They’re offering podcasters a way to understand their audiences and —finally —to make some money. These new platforms deal with the holes in RSS by, well, getting rid of it all together. Run a search at tech business portal Techcrunch for “new podcast platform” and you’ll get a list of a dozen new and evolving businesses —ACast, Megaphone, Anchor, BuzzSprout, Stitcher… The business models vary but the pitch is obvious. In a hundred boardrooms, this week, someone will say— “audio is the new video, something, something, true crime podcasts are like crack for millennials, something, something, there’s no Netflix for audio, something, something, the targeted ad opportunity is enormous, something, something…”

So these platforms are going to start acting like platforms. Expect them to fight it out for ownership of the big podcast brands —just like Amazon and Netflix do with the TV brands. The audio business is inevitably going to look more like the crazy mosaic of rights, territories and exclusive deals of the video business. Your latest binge listen will be on ACast or Apple Podcasts or Stitcher but not “wherever you get your podcasts.” And the abandonment of cuddly old, simple old RSS will result in a wave of new and more intrusive ad formats. Get ready for personalised ads, ads that know where you are, ads that follow you from platform to platform and —you’ll like this —ads you can’t skip.

Common sense suggests that even the fancy end of audio is never going to be worth more than a decent fraction of the video streaming business but the investors diving into podcasting now are serious about it and want to build a business with real scale. And, let’s be honest, it’s in everyone’s interest for them to succeed —the mattress ads are not going to sustain a heterodox commercial audio ecology for long and the planet’s already stretched public service providers certainly can’t do it on their own.

Realism (and good evidence from the rest of digital media) also tells us that although we’d love to see a hundred flowers bloom, the audio business is going to pretty quickly shake out to a small handful of big platforms —with one global player making the rules, setting prices, shaping supply.

Can we take an educated guess as to who that will be, though? Which of the handful of serious-looking platforms will dominate podcasting?

So who’s going to win podcasting?

All right, I’m just going to say it: it’s going to be Amazon (I’m taking bets —leave a comment if you want some of the action). “But hold on”, I hear you say, “Amazon’s audio platform is called Audible. It’s a repurposed audiobook catalogue and, despite years of effort, they haven’t made even a tiny dent on the non-audiobook business. An audio outsider like Amazon doesn’t stand a chance!”

You’re right, of course. They’re slow off the mark and Audible itself is a pretty charmless environment in which to win over the next generation of audio nuts. There’s even a reasonable chance that Amazon will just ignore the podcast revolution all together, write the whole thing off and focus on the Billions they make from retail, video and web services. But I doubt it. Here’s why it’ll be Amazon that wins podcasting:

They’re everywhere. Amazon apps —shopping, video, music, Kindle, Alexa —have hundreds of millions of installs between them. Installed on audio-ready devices and pre-loaded with credit card numbers and a detailed purchase history.

They have a billing relationship with millions of people. There are around 90M Amazon Prime accounts in the US and it’s estimated that each spends $1,300/year with Amazon (about a third of UK households are already paying members). Few have even a fraction of Amazon’s customer base, let alone paying customers. Audible itself is a subscription platform that’s cross-promoted to Prime subscribers and Amazon has made clever use of cross-promotion in launching a dozen other services over the years.

They know how to sell media to millennials. They’re no Netflix but Amazon Prime Video already reaches millions of people in podcasting’s demographic sweet spot. Audible, in fact, already commissions a small number of ‘original audio series’ (they don’t call them podcasts, of course) and cross-promoting them to the company’s other audiences will effortlessly produce the kind of listening figures that most podcasters would kill for.

They own the most promising new interface to the global store of audio. Echo is the best of the voice-controlled devices and it’s already in 11% of US homes (an adoption rate that looks similar to that of radio in the 1920s). Better yet, it turns out that audio is one of the most popular services among users (according to this RadioPlayer research, from the UK, radio is the most popular category of audio on ‘smart speakers’). The affordances are perfectly aligned. Who knew?

Their platform will permit all the scary ad-tech. No dependence on sad old RSS here so there’s nothing stopping Amazon from dynamically serving you ads, right into the audio stream —and all targeted with insane precision, because…

…let’s face it, they know everything. It’s not clear exactly what information the Audible app captures as you listen but it’s safe to assume that —in addition to Amazon’s detailed understanding of what you like to buy— they know where and when you’re listening, how often you listen, how far you get through a download, how many sessions it takes you to complete a listen. And it would be trivially easy for Audible to capture a more detailed picture, using the sensors in your smartphone to learn what you listen to while you’re working out, while you’re commuting, on your lunch break, in the woods, on an aeroplane, in bed, in the vicinity of a military base…

This level of detail —especially if passed on to producers— would profoundly change podcasting, becoming the primary influence on choice of formats, voices, themes and styles across the business —just as it has in journalism and video. More, it might secure commercial viability for an industry that, frankly, still looks shaky.

The only question, for Amazon, is going to be “how much is this worth?” If podcasting turns out to be worth a tenth of video streaming, it might be worth the engineering and marketing effort to relaunch Audible as a podcasting platform. If it’s a hundredth, they’ll hesitate and may not even bother. And, incidentally, this may explain why Amazon has been slow to capitalise on Audible’s ubiquity. Is Amazon holding back because their data tells them that podcasting’s just not worth it? I do hope not.

Bonus episode

I’m aware that I’ve left out some fairly important platforms. Not because I don’t think they’re important, although, actually, I don’t think they are quite as important. But I accept that some of these stories are going to be as interesting and influential as the big ones above…

Spotify’s Podcast offer is pretty rich but it sits alongside music uncomfortably —and there are some pretty hideous interface issues. The Swedes are about to raise a lot of money via an unusual direct listing on the NYSE, though and they could choose to blow a chunk of that money expanding their catalogue of commissioned audio content (and tidying up the UI). That would be interesting in itself, because it would represent an opportunity for institutional and retail investors to get involved in the next generation of podcasting platforms pretty directly.

Meanwhile, Who We Be, a new podcast tied to one of the platform’s big urban music playlists and presented by British DJ and broadcaster DJ Semtex, is probably a model for the next wave of speech audio from Spotify.

Soundcloud is also a podcasting company. Their stock of audio is enormous and the “wherever you get your podcasts” model means that lots of producers publish their stuff on Soundcloud automatically. But the money Soundcloud raised in 2017 is going to be devoted to keeping the company afloat and finding new focus. Building a new podcast brand is off the to-do list.

Google’s approach to podcasting is to make it another checkbox on the long list of services you can access via the mobile app. This checkbox approach commodifies the business, though, and makes it difficult to build a big, prominent brand. But we probably shouldn’t rule them out: they’ve already commissioned at least one original podcast series (although it seems to be on hiatus).

Apple’s podcast app got better in 2017. The data available to producers also improved. And no other platform can match the support the company’s podcast teams supply to producers in big markets. The charts and category pages they curate remain the primary way of finding the good stuff for listeners. There’s an obvious opportunity for Apple to convert this epic organic advantage (and a tiny shred of that enormous cash pile) into a serious new commissioning platform —although the company’s careful progress into video commissioning is probably a good guide to how this will unfold (Apple has commissioned eight TV series since hiring Jay Hunt in October last year, though).

Smartphone apps are very interesting (they’ll need a post of their own, really). They’re interesting because they can deliver a lot of the important platform benefits —a logged-in experience, dense listening data, recommendation and discovery, even a billing relationship —while piggy-backing the existing RSS-and-string-based infrastructure. No need to reengineer the distribution layer. These apps depend on the continued availability of content via the old RSS infrastructure, though, so if the big podcasts disappear into walled gardens, their catalogues will shrink and they’ll be left with only the DIY end of the market.

The podcast category in your app store is packed with interesting apps —some of which have been there for years and some of which promise an improved experience for listeners and even some income for producers (if you’re using one of these apps and you like it, leave a comment —I’d love to know which apps are doing this right).

The podcasting boom will support a complex services ecology. Lots of smaller businesses —usually with a national or niche focus —are providing services to the flourishing DIY end of the business. Hosting, audience measurement, editing, audio optimisation and —in some cases —advertising/sponsorship services. This is a good sign —a complex services ecology = a viable industry. Some of these businesses will try to scale up to become platforms. One or two might actually achieve it.

The BBC and other big broadcasters and publishers are in a tricky position. They feel an obligation to participate in the audio explosion. Some, of course, are grizzled pioneers in this business —try to count the number of online audio products launched by NPR, The Guardian and the BBC over the years and you’ll soon run out of fingers. They have big existing audiences, production talent and engineering know-how to bring to the game but they can’t easily challenge the mega-platforms on range or promotional clout. This is going to get interesting.

The first of this series of posts about podcasting is about the creative battle. Also read my 11 essentials for the modern podcast.

Picture by Gauthier Delecroix, on Flickr.

Podcasting—welcome to the symphonic era

This is not about the 90% of podcasts that are still three people at a table talking about something. Nor is it about all those podcasts that are basically a byproduct of radio production. It’s about the new stuff—the bigger, glossier, narrative formats that are going to change audio and storytelling for good. This is part one of a two-part series. Part two is about the platform battle and you should also read my 11 essentials for the modern podcast.

Podcasting is evolving fast. There’s a strong sense that we’ve passed some kind of tipping point, that this is how we’re going to consume audio (the stuff that isn’t live radio or music anyway) from now on. The creative battle is under way—formats, genres and production norms are all in play. The platform battle—who will distribute this stuff, who will own the payment gateway and the relationship with the customer—is about to begin (more about that bit in my next post).

Storytelling formats are going to set the tone and dominate the fancy end of podcasting. The iconic 2018 podcast is going to be a true story told using the techniques of fiction. Netflix is our model here, not a radio station. And things are moving fast. Serial —the groundbreaking hyper-addictive episodic story that started all this—already sounds old-fashioned, under-powered. But these formats are expensive and there’s a premium on scarce production talent so only well-funded organisations can play. That means it’ll probably be the three-letter incumbents (BBC, NPR, ABC…) and the newer, specialist outfits with their own funding (Panoply, Gimlet, Radiotopia…). The specialists have a significant head start.

Production is rich and multi-layered. This is what I mean by ‘symphonic’. In this podcast category, where Heavyweight, Mogul, Slow Burn and maybe a dozen other big shows live, the pressure to create complex productions —and emotionally rich stories —is going to be enormous. One inspired amateur with a microphone this is not. Resources and talent will be coordinated, teams assembled, walls filled with Post-It Notes. The best of the new generation of podcasts are already made on a pretty grand scale —they’re big productions with credits to match (16 people for this episode of Mogul, 47 for scripted drama Bronzeville —including a cast of 18, a caterer, a historic consultant and two executive assistants). Some of these podcasts are big enough and confident enough to have a ‘making of…’ episode.

There’s a tone of voice, a recognisable tenor, to these bigger, more ambitious podcast stories. Keywords: warm, humane, emotional, generous, personal, authentic. Podcasters like Jonathan Goldstein. Manoush Zomorodi, Roman Mars, Helen Zaltzman (and the form’s honorary Godfather Ira Glass, natch)… have a thoughtful, subjective, ironic way with their material. And in the writing there’s a deliberate continuity with the tradition of serious, crafted, non-fiction storytelling that produced all that amazing 20th century writing—the New Journalism, the whole clever lineage of long-form magazine writing too. Joan Didion in New York Times Magazine, Hitchens in Vanity Fair…

There’s a ‘big city’, Public Radio, New Yorker feel to this stuff. Nothing rushed or half-baked about symphonic podcasting. This is luxury storytelling for nice people who probably still buy the Sunday papers. Audio that flatters the listener’s intellect and is as likely to make you cry as to smile. Incidentally, of course, all of the symphonic pioneers are American. This is not because they’re any cleverer than the rest of us, but principally because an economic model—venture capital—exists there that can mobilise large amounts of money for speculative productions that may never break even. Everywhere else producers are stretching existing production budgets or bootstrapping like mad. Radio producers, who think they already know all about audio storytelling, are going to have to learn some humility, too. Their skills will be vital but their cottage industry economics won’t.

TBH these formats can sometimes be a bit sickly. If Jonathan Goldstein makes me cry in the first reel again I’m going to unsubcribe; the enveloping sound world of Jad Abumrad’s gripping Supreme Court documentary series More Perfect is so detailed and so rich as to be a little too much. Everything in high-end podcast land is amped-up, slightly overdone. Look over the shoulder of a producer in this part of the market and you’ll see a workstation with dozens of active tracks. There’ll be subtle and engaging sound design, a commissioned score and incidental music and post-production effects. You can almost hear the producer’s titanic effort to fully engage the audience’s feelings. The signature emotional tone of the symphonic era is slightly over-wrought. Or maybe I’m just being too British about this. Anyway, dial it down, gang. No need to lay it on so thick. We can feel it.

Of course, I don’t want to over-do the analysis. Is podcasting evolving into a new and influential journalistic form, with its own shape, its own creative logic and its own economics? Definitely. Will it become grand and influential, will the symphonic era produce a generation of famous voices, writers and producers? Will it shape the culture? Possibly.

So that’s the creative battle. My next is about the platform battle that’s about to begin—who will distribute the new generation of podcasts, who will own the customer relationship and who will make all the money. And further down on this blog, you’ll also find 11 essentials for the modern podcast. Meanwhile, like I said, I think that we listeners are going to spend a lot more time sobbing into our lattes in the symphonic era, so here are:

Five episodes from symphonic era podcasts that will make you cry.

Heavyweight —Isabel (Gimlet). I love Jonathan Goldstein’s Heavyweight. I think it’s the signature symphonic podcast. Even though I’m frankly ashamed of the way he always makes me cry (and usually when I’m on my bike, making it all very inconvenient, not to say dangerous). This one’s no exception. Goldstein provides evidence that he can achieve an emotionally complete storytelling experience even when his main character refuses to provide the resolution we all want.

The Allusionist —Joins (Radiotopia/PRX). One of my absolute favourite podcasts, Helen Zaltzman’s The Allusionist is all about language, defined very broadly. She’s evolving the show in a really interesting, dare I say, symphonic direction. This episode is an essentially un-presented sequence of very moving voices from the trans and non-binary community and it will also make you cry).

Ear Hustle —Left Behind (Radiotopia/PRX). This podcast, which came to the world via Radiotopia’s talent search Podquest (more about that here) is a stunning piece of work —my podcast of the year, by a mile. And one of those rare productions —in any medium —that I think stands a chance of producing actual social change. It’s made inside California’s San Quentin Prison. This episode, like I said, will make you cry.

Mogul —August 30, 2012 (Gimlet). Gimlet’s lovingly-made six-part doc about hip hop impresario Chris Lighty. This one (which is the climactic episode, so SPOILERS) will also make you cry. Sorry.

Note to Self —You Deserve to Die (WNYC). Manoush Zomorodi’s show is not the kind of podcast that would normally make you cry. It’s a podcast about the way we live in the networked era and episodes are usually on a spectrum that goes from self help to consumer advice to WTF-is-Silicon-Valley-doing-to-my-brain? This one is going to make you cry, though. Seriously. (and it’s the only one on this list that comes from an actual radio station— WNYC).

This post appeared first at Medium.com. Part two of this series is about the platform battle. Also read my 11 essentials for the modern podcast.

11 essentials for the modern podcast

This is one of a short series of posts about the evolution of podcasting. The first one’s about the new wave of ambitious, highly-produced storytelling formats – I’m calling it ‘the symphonic era‘ and the second is about the epic platform battle that has just begun. So I thought I’d collect some of the basic elements of the ‘symphonic’ podcast here. Consider this an incomplete list of things you need to do.

Do bold, generous promotion. Give whole episodes to friendly podcasts and encourage them to tear them down and re-edit them for their own purposes – This American Life uses a cut-down ep from Gimlet’s ‘Heavyweight’. Radiolab carves up a whole ep from NPR’s ‘Rough Translation’. Edits can be really radical, an episode can be totally remade and feel very different but this is great promotion and very flattering to the source.
Tease cleverly. Publish a ‘season preview’ or an ‘episode 0’ ahead of the main series. Heavyweight just did this and it really builds excitement. This would also work for returning on-air podcasts from broadcasters – i.e. episode 0 would be online-only, so could have a different tone and maybe a looser format and throw forward to ep 1.
Publish a ‘making of’ episode (and a blooper reel and a cast interview and a story follow-up etc. etc.). Major productions like Gimlet’s ‘Mogul‘ and ‘Bronzeville‘ have done this – wringing the maximum possible value from their expensively-created content.
Commission music. Whatever your podcast is, whatever the theme, no matter how unnecessary music may seem to your theme or format. It will amp up your podcast, make it feel more grown-up, more symphonic. I love the clever, lightweight music they use on The Daily, for instance.
Mine the archive – and other people’s archives. You’ll need permission but, if it’s there, this is essentially free content. 99% Invisible resurfaces old Public Radio episodes that happen to fit a current theme. Radiolab routinely fills gaps with older eps, minimally reworked or updated.
Invent formats – and give them funky names. Like Mogul’s ‘Cameos‘ – mini-episodes between the main ones that don’t carry the story.
Oh, and do mini episodes between the big ones. Minimal effort, possibly built from unused tape from the main eps. Be cheeky about this, don’t feel you always need to create original content, don’t be uptight about your publishing schedule. People will be excited when they see an unexpected ep land.
Put on live events – it turns out this will work with literally any podcast. Seriously. It will add energy, provide material, excite contributors and suggest new approaches. Does Sawbones, the ‘marital tour of misguided medicine’ need live shows in venues all over the USA? Not really. Does it work? Yes it does. In the US now, wherever you live, your local theatre or live venue will definitely have at least one live podcast show in the schedule. It’s the rules.
Find a way to include the voices of listeners – even if you just get them to read the credits. The NPR Politics podcast gets people to read out the disclaimer about the podcast probably being out of date by the time you hear it (they call them ‘timestamps‘).
Do ’emergency episodes’ – and not just for news podcasts. Any time there’s a real world event to respond to, get into a studio and lay down 20 minutes of chat. It connects you with the news, makes you seem up-to-date. Here’s one from the excellent FiveThirtyEight Politics podcast.
Provide credits – name everyone! It gives your podcast weight, makes you look like a player. And give people titles. Anyone who ever listened to a recording is a producer, anyone who ever advised you is an exec.

Ian McMillan’s eight favourite podcasts

Radio 3’s Ian McMillan was on a special edition of the Radio Today podcast all about the station the other day. Turns out he’s a connoisseur of the podcast form. He gave Trevor Dann a list of his favourites:

the various Monocle podcasts, especially Tyler Brûlé’s books and magazines podcast The Stack, The Urbanist and The Menu.
The University of Rochester’s 3% – books in Translation.
The Bad at Sports contemporary art podcast.
The All Things Radio podcast, an American radio industry bulletin.
The Radio Today podcast, natch.
The Radio Stuff podcast.
The Guardian’s venerable industry podcast MediaTalk.
The Freelance Web podcast, which is for people who make their living as… well… freelancers on the web.

BTW, listen to the end of the Radio Today podcast and you’ll hear Radio 3’s head of speech Matthew Dodd and Falling Tree‘s Alan Hall talking about doing speech on a classical station and Between the Ears‘ twentieth anniversary.

My inspiration

Front cover of Byte Magazine for February 1984, featuring a picture of the new Macintosh computer — Front cover of Byte Magazine for February 1984

I’m not a geek. I missed the boat. When I left school they’d just acquired a computer. It was a mysterious, chattering presence in a room in the maths department – a teletype connected to a mainframe somewhere – and I never met it.

But when I first encountered a computer – in a roomful of brand new Macs at the Polytechnic of Central London in 1985 – and set about learning about them, I beetled off to one of those Soho newsagents that still set my heart racing, with their rows and rows of thrilling imported titles, and looked in the computer section. The magazine I settled on and made my bible was called Byte – the small systems journal.

Byte was recklessly terminated in 1998. I still miss it. It was a quite awesome monthly crash course in IT – a kind of undergraduate degree in magazine form. Long, gripping articles about chip design, network architecture, software and AI. I owe more to Byte than to any other source of knowledge about computers.

Byte, which for most of its life was published by McGraw Hill, was no web pioneer. In fact, for a while, during all the really early frenzy (during which I helped publish a magazine that was all about the web), Byte was almost a holiday from the Internet – a place you could go to read about VLSI chips and ethernet while the rest of the world was going web crazy. When they decided to have a go, they did it in a very Byte way, though.

They put a man called Jon Udell on the case – he was a staff writer and he was given the job of building the magazine’s web presence and documenting the process month-by-month for readers. He brought the whole thing to life with a really forensic attitude to the emerging tools – and invented a bunch of new ones along the way. These days he works at Microsoft and he’s an influential geek with an interest in all sorts of developing areas – and his ‘interviews with innovators’ are published as part of the IT Conversations podcast.

But this one’s a bit different – a rather modest, one-hour conference speech about ‘the architecture of context’, in which he lays out his own, partial history of the net and remembers some of the lessons he learnt in the Byte days. Fascinating and inspiring.

[Sad to say the MP3 mentioned below seems to have gone]

The MP3 is from the IT Conversations podcast. Definitely worth signing up.