James Cridland's blog

A radio futurologist writing about what happens when radio and new platforms collide

« | Blog index | »

The P&M Awards – a review of a preview

Posted on Thursday, October 22nd, 2009 at 11:36am. #

Tonight, as a judge and trustee, I’m off to The Radio Academy’s Promotion & Marketing Awards. It’s at a place I’ve never been before in Great Portland Street, and I’m hoping to bump into a lot of old friends and have a good evening. While the Sony Radio Academy Awards are good, it’s always good to see some of the rest of the craft in radio in more specialised awards, and I’m looking forward to discovering some neat ideas.

However, this isn’t actually a blog posting about the P&M Awards. It’s a blog posting about a blog posting about the P&M Awards, a kind of recursive blog-will-eat-itself posting, much like the BBC’s inevitable Editor’s Blog posting about the BBC news coverage of the BBC Trust decision about whether the BBC Question Time discussion with the BNP’s Nick Griffin should go ahead. (Oooh, topical. Excellent. Come, Google spider, come.)

The blog posting I’m typing about is Steve Martin’s interview with Trevor Dann; and it’s very interesting for two reasons.

It’s out before the event.
It would have been easy for Steve to take an audio recorder to the P&Ms, and interview Trevor; publishing it the next day. However, it would have been lost in the noise of the award winners, and would have obviously included chats about the winners and nobody else. By publishing before the event, it builds excitement and buzz about the event, and will draw traffic because it clearly contains clues as to who has won and who hasn’t. Maybe. It’s probably no surprise, then, that radio news websites have picked up on it.

It has a transcript
This is the main brilliance of this post.
1. Google can’t index your audio. It can’t. Trevor’s mention of Radio City’s “Make My Day” feature would not have appeared on Google if Steve had only posted the audio. Yet, Google can index text. And does. And has.
2. I can read faster than real-time. So, I was able to flick through this interview without, bless him, listening to Trevor’s voice, discovering what he was talking about, in a much more efficient way than simply listening to the audio.
3. It’s surprisingly cheap to do. Transcribing a ten-minute interview in this way looks as if it might have cost around £20, judging from a quick Google search. If you want to put a little more work in, it could be even cheaper. This is not outside your price range; and it’s something that can easily be outsourced to India or similar areas.

What surprises me, therefore, is why more radio stations don’t transcribe their interviews. While I enjoyed the Chris Evans interview yesterday at Absolute Radio, or the programme called The Interview on the BBC World Service, why not transcribe these, too, to add to their Google-ability? And why do we treat our audience’s time so poorly, expecting them to sit through twenty minutes of audio instead of ten minutes of speed-reading? Is that good user-experience?

As ever, your thoughts are welcome. Incidentally, you’ll now see feedback from Twitter, Digg, and a few other places in my comments, courtesy of a new plugin I’ve discovered. Hopefully this’ll enable you to discover new people to follow on Twitter, too.

7 comments

Tim Page
commenting at October 22nd, 2009 at 11:57am

Why don’t more sites use audio searches – ie actually crawling the recording phonetically for search terms?

Although it can never be foolproof and I guess might disappoint the non-techy user who’s used to Google, I’m always impressed by the Audionix system which works on the Beeb’s internal audio logging system AutoROT.

Given Google’s reasonable accuracy interpreting my search wishes on iPhone, it can’t be far off!

Steve Martin
commenting at October 22nd, 2009 at 12:11pm

Glad you enjoyed it James. Building excitement before the event is second nature to promos folk of course. It wasn’t calculated, but just felt right.

The transcription cost me £12.50 plus VAT and was outsourced to, er, Richmond upon Thames. There’s a link to the company on the blog post at http://earshot.tvi.gg

I have little doubt that Google listening robots with digital ears will automate all this one day.

Adam Bowie
commenting at October 22nd, 2009 at 1:22pm

To answer Tim Page, it’s probably the expense of the thing.

To be fair to our own press team at Absolute Radio, we do hire transcribers for major events like V Festival where lots of interviews are being carried out, and we think we might get good press out of it. These transcribed interviews are then passed on to the usual media outlets.

But it’s also fair to say that access to these tends to be limited to press, and when the same interviews appear on our website, the transcription is not placed alongside which I think is a fair comment.

Drew
commenting at October 22nd, 2009 at 2:10pm

A nice reminder, James and with the apparent trend of written interviews also publishing the original audio/video, radio and podcasting shouldn’t be left behind.

I strongly feel one should only provide transcriptions that are accurate and verified (preferably by the interviewer), otherwise you may be misquoting or even libelling the interviewee. Regional accents can make “can’t” sound like “can” and failing to annotate a response as being sarcastic can confuse the whole issue. In both cases you’d be publishing the exact opposite of what was meant.

A bad choice of transcription company (unfamiliar with cultural references or the technicalities of the subject matter) will just lead to more question marks, more corrections and the risk of overlooking something that sounded right – but wasn’t.

This isn’t just about being able to find the audio using text search – it’s about representing exactly what was said and becomes extra important since those quoting the interview would likely do so using the text version and not even listen to the original audio.

For those of us from a radio background the references have always “spoken for themselves” and, unlike print journos, we didn’t need to confirm, note down (or at least research later) spellings & details for every reference cited – be it a person’s name, little-known band, unfamiliar website or technical term.

I personally hope Google isn’t planning to index audio just yet – far too many people trust Google Translate which, for example, still translates the Dutch word for twenty (twintig) as “eight p.m.” making a mess of many news stories on one site I read! So audio automagically appearing in searches could make a real mess of a news interview and damage your reputation.. best DIY ;)

Francois
commenting at October 22nd, 2009 at 5:02pm

Interesting post James. I wrote something on the same line some time ago: “Podcast Transcripts and the Mechanical Turk” bit.ly/2F6r2M

gavin
commenting at October 22nd, 2009 at 11:48pm

Here in Australia, the ABC transcribes lots of TV and radio reports and puts them up on the website along with the audio e.g. the entire daily content of both AM and PH can be found there http://www.abc.net.au/pm/

NPR – doing everything right - blog - James Cridland
commenting at November 24th, 2009 at 10:28am

[...] does extensive transcripts I’ve written before about the tremendous benefit of doing transcripts: their system involves a human checking automated output and publishing it [...]

Leave a comment

Here's my commenting policy