Way back in 1979, before YouTube’s three founders could either walk or talk, The Buggles claimed that “Video Killed the Radio Star.” Queen countered in 1984 with Radio Ga Ga, in which Freddie Mercury sang (of radio);
We watch the shows, we watch the stars
On videos for hours and hours
We hardly need to use our ears
How music changes through the years.
Let’s hope you never leave old friend
Like all good things on you we depend
So stick around cause we might miss you
When we grow tired of all this visual
You had your time you had the power
You’ve yet to have your finest hour
As many of you know, I’ve been running a poll over the past week, to understand whether my future podcasts should remain audio-only (my not-very-well-disguised preference), or start to include more video. And, at least for the type of content I create, it would appear that the lyrics by Queen’s Roger Taylor are spot on; audio wins.
There’s some more detail to tease out of the results, after the break…
Seventeen individuals completed the survey over a period of seven days. Asked “Which type of podcast would you like me to produce?” (the only mandatory question), 15 selected ‘Audio only,’ 2 selected ‘Both audio and video,’ and none selected ‘Video only.’ Amongst the (optional) comments, there was a recognition that video would have value if there were more to watch than just two talking heads. I don’t see any point peppering a video with screenshots of a web page that the viewer could just visit for themselves, but there is certainly value in showing people how to complete some task or process. Luckily ( ) my podcasts don’t involve that degree of minutiae…
“A video version would be a little more trouble for us, a lot more for you, and for what? To watch badly-lit panelists stare into their webcams? Stick with what you have. It’s great.”
The remaining questions were optional, but most respondents chose to answer them.
Asked ‘How often do you listen to my podcasts,” 50% reported listening to “all” or “most of them,” 30% had heard “one or two,” and 18% had not listened to any.
Most of my podcasts tend to run for between 40 minutes and an hour. Asked about their preferences for length, 70% of respondents stated a preference for the current length. There were, of course, a number of caveats in the comments. These broadly suggested that a podcast should be as long as it needed to be, rather than padding a conversation in order to reach some arbitrary length. That’s pretty much what I do just now, so that’s good.
When I started podcasting, I used to slave over the audio for hours. I edited out every ‘um,’ every ‘ah,’ every plosive, and every pause for reflection. Now I don’t bother. Instead, I add a short intro and edit out the worst of the howlers (coughing fits, retracted statements that might make lawyers twitch, etc). From my perspective, that gives the best balance between editing time and listening quality, and the survey’s respondents agreed. One respondent said I shouldn’t bother doing anything at all to the audio file, 12 said “do what you do just now,” 2 asked that I “do more to edit the ums and ahs,” and one opted for “edit to the max – give me the sound bites, and voice overs and summaries.” I think I’ll quietly ignore that last one, and keep going as I am.
So… keep doing audio, keep them about the current length, and stick with the current level of editing. Isn’t it great when a survey tells you what you want to hear?
That said, there may be scope for doing some talking heads-type video at events. That sort of video is about capturing atmosphere as well as getting the interview… and would probably be much shorter. 5-10 minutes, maybe. We’ll see…
Disclaimer: lots of wordy verbiage, crammed full of caveats, hedges, and warnings about sample size, survey methodology, and bias.
And as for the songs? Well, judge for yourself. The Buggles video off YouTube doesn’t appear to want to embed (who knew that WordPress had a taste filter?), but here’s Queen…
- A conversation with Richard Wallis, an experiment, and a survey (cloudofdata.com)
(Cross-posted @ Paul Miller – The Cloud of Data)