Questionable podcasting stats from Pew Internet

In show #21 of The Hobson & Holtz Report on Monday, Shel and I discussed research on podcasting from the Pew Internet and American Life Project that was released on that day. Pew’s research information we discussed included this statement:

Some 29% of the 22 million people who own iPods/MP3 players have downloaded podcasts. That represents more than 6 million people.

It now appears that this information is not what it seems.

According to a Techdirt report on Tuesday:

[…] the research director behind the study clarifies (after the fact, of course) that the study actually asked people: "if they had ever downloaded a podcast or radio Internet program." So, out of 200 people [who were interviewed], they got 60 to admit that they had maybe at some point downloaded an internet radio program (which is not necessarily the same thing as podcasting) — and from that they put out a report with the headline that "6 million American adults have listened to podcasts."

If that’s true – and there’s no comment on the Pew website – then I’d have to question the validity of much of Pew’s statistics. Disappointing to learn of this.

However, while trust in Pew has taken a bit of a dent, I don’t think this news diminishes by any means what clearly is a growing trend. I feel pretty comfortable saying that and relying on data such as the numbers of podcasts that are catalogued in the various podcast directories like Podcast Alley, which have more and more every time you look.

There’s also reliable data from FeedBurner in February that measured the growing number of podcast RSS feeds that their SmartCast service handles (see my post and FeedBurner’s commentary).

These are plenty of valid indicators.

Update 16 April: Just saw, via a link on Adam Curry’s blog, that Pew Internet stand by their data. See their post for a detailed explanation.

5 thoughts on “Questionable podcasting stats from Pew Internet

  1. I think they have got their definitions wrong. If I download the Today programme on BBC Radio 4, I have downloaded a broadcast. If I subscribe to any piece of audio designed to be played on a desktop or portable player, you can argue I am picking up a podcast. Sounds like the same sort of confusion as when people were asked if they listened to “cable radio”. 80% in Holland said yes. Turned out they had a radio – and it had a cable on it – the power cable. It wasn’t connected to the cable at all.
    But are there real stats coming out about Podcasting? Yes. NASA has published figures. IT Conversations is going open source.
    Many podcasters are vague when it comes to mentioning audience figures, or downloads-per-show. I think it is in the region of 30,000 downloads per show for the popular ones, down to just a few dozen for most of them. Some people seem to forget that you are only as good as your last show…if it bites….so will the figures. So how is NASA doing? Between its December launch and the end of February 2005, listeners accessed the Science@NASA podcast more than 360,000 times. That compares well to subscriptions to science magazines in the store (between 15,000 and 60,000 readers depending on the country).
    I disagree with the remark on FIR that audio quality and presentation don’t really matter, so long as the content is good. Whilst I think that authentic voices clearly have an edge on the over-polished radio presenters, I do appreciate someone who has taken the time to craft their podcast…too much of the “slice of my life, uncut and unedited” is as exciting as watching paint dry.

  2. Podcasts are just mp3 files enclosed in RSS feeds. I can subscribe to an RSS feed and have a program download any mp3 linked to in the post. Maybe I want to just subscribe to RSS feeds of mp3 blogs and download the mp3s at my own choosing. RSS feeds are just a kind of XML and XHTML is a reformulation of HTML as XML. I can have a program check an XHTML page and download all linked mp3s. Hey, I could even be enclosing ogg vorbis or *shudder* realmedia files in my RSS feed.
    Clearly, defining podcasting so narrowly doesn’t prove to be all that useful. Fussing over what’s a podcast, what’s a weblog post with an MP3 linked only inside it, or even attached to a node as Drupal does, is splitting hairs. The Pew study speaks to the popularity of time-shifted, citizen created radio – whatever you want to call it.

  3. Jonathan, I agree with you on the definitions point. There is quite a bit of mis-communication, let’s say, surrounding podcasts and what they are. As Will says, they’re MP3 (typically) files enclosed in RSS feeds. So the name ‘podcast’ is to do with the distribution, not the medium itself.
    However, this is one of those things where it looks like a losing battle to apply a purist approach to closely defining what this is. People will talk about podcasts when they mean going to a site and downloading a file, as in “you can download my podcast from…”. But so what?
    I think your point on audience figures is more interesting, and is one that will be increasingly important for some podcasters who will need to show evidence of the reach and influence their shows have, especially if things move along into the commercial area where sponsors get interested. As with any communication medium, this enters the traditional area of measurement and analysis. In the case of the show Shel and I do, we’re becoming very conscious of this need. Measuring only by bandwidth consumption just isn’t good enough, so we’re looking at some methodologies that will give us more credible analysis. All we can do right now is work out averages. So, for example, we know from bandwidth stats that our show currently attracts anywhere between 500 and 1,000 downloads (meaning both manual downloading from the podcast site as well as retrieval via the RSS feed) per show. To put that in context, the average for the first month we started (January) was about 30 to 50.
    What I’d really like to know, though, is what percentage of those numbers translate into actual listeners. We don’t know that yet.
    The numbers you mention (30,000) apply to those shows that have wide general appeal. Adam Curry, for instance. Or Dawn & Drew. The great thing about this medium – it’s nanopublishing, to grab that vogue word – is that the production costs are relatively minute which gives literally anyone an opportunity to start a show where if only a couple of hundred people like it, that’s still good enough to justify doing it.
    And on audio quality and presentation, well, I’d have to agree with you there – they do matter. Yet it’s a subjective point, isn’t it? What may be as exciting as watching paint dry to you, could be the most stimulating thing to someone else.
    That’s another great thing about this niche medium – there’s something for everyone’s tastes!
    Shel posted a detailed commentary on his blog on Monday about podcasting and Darren’s post. Worth a read, especially the comments –

Comments are closed.