Enjoy great online radio at AccuRadio

Internet radio larger than yesterday's press reports

Posted on: 10/18/2007


How popular is Internet radio as a medium today?

It’s a lot larger than one would imagine based on reports that appeared in radio trade publications yesterday (e.g., from Inside Radio: “Internet radio: Potential is great, listeners are few. Spring 2007 data shows less than 1% of diaries contain listening to Internet radio”), but the exact answer is by no means clear.

The flawed coverage yesterday was due to a flawed PR release issued by Arbitron on Friday, which used the “less than 1% of diaries” statistic but failed to mention that the statistic ignored listening both to Internet-only radio stations (e.g., Yahoo! LAUNCHcast and Pandora) and to exact Internet simulcasts of AM/FM stations (i.e., simulcasts without ad insertion).

The larger problem here is that Arbitron has three or four different data points regarding Internet radio listening, and they don’t jibe with one another:


Per the methodology by which Arbitron edits and tabulates diaries, listening to Internet-only stations (e.g., Yahoo! LAUNCHcast and Pandora) is thrown out entirely (just as they would ignore diary entries like “‘Heroes’ (Channel 5)” or “I went to see ‘Pirates of the Caribbean’ at the mall” or “Ate a Southwest Salad at McDonald’s”). As for listening to exact Internet simulcasts of AM/FM stations (i.e., with no ad insertion involved), it’s simply credited as listening to the AM/FM station.

What Arbitron’s PR people meant to say on Friday is that less than 1% of diaries show mentions of listening to Internet simulcasts of AM/FM stations with ad insertion. Assuming about 250 million Americans age 12+, that’s less than 2.5 million people listening to that specific form of Internet radio each week.

One problem with using diaries to answer this question is that the design of the diary (see illustration) asks diarykeepers to tick either an “AM” or an “FM” box for each listening event.

Notwithstanding what the diary might say in the instructions on other pages, a reasonable person might conclude that they are only supposed to write in AM and FM listening —and not the TV programs they watched, the movies they saw, or the Internet audio they consumed.

ComScore Arbitron

According to results from a 200,000-person panel maintained by the comScore research organization and used by a joint venture between comScore and Arbitron, about 4.7 million people per week listen to some combination of AOL Radio, Yahoo! LAUNCHcast, Live 365, Got Radio, 181.fm, and Big R Radio (the latter three appearing in the monthly reports as Ronning-Lipset Radio’s “RL Select” network). (Oddly, the comScore Arbitron results say that there are no consumers who listen to more than one of those properties in the same week. This must be an error; we’ll address this in a future isse of RAIN.)

In addition, comScore Arbitron reports that about 1.4 million people per week listen to at least one simulcast of a Clear Channel AM/FM station. That’s a total of about 6.1 million people per week listening to one of those seven Arbitron clients.

Adding in reasonable estimates for the Internet audiences of CBS, Entercom, Bonneville, Cox, Citadel, and other broadcasters, plus the audiences of the properties that subscribe to the other Internet radio ratings service (AndoMedia’s Webcast Metrics), plus the audiences of properties like Pandora that subscribe to neither service, one might reasonably come up with an estimate in the 10 million to 12 million range.

Edison/Arbitron Studies

Each January since January 2000, Edison Media Research has conducted follow-up telephone interviews with about 1,800 former diarykeepers regarding their new media listening habits. This year’s study (January 2007) indicated that about 11% of P12+ Americans — i.e., about 28 million people — would have said that they listened to Internet radio in the prior week.

(Incidentally, this 11% statistic was down from 12% in January 2006, but a study with that sample size for a finding of this size has about a 2% margin of error, so it’s reasonable to guess that 2006’s “12%” finding was a little higher than reality and 2007’s “11%” might have been a little lower than reality.)

Personal People Meter (PPM)

There’s a fourth data source that Arbitron has access to but I believe isnot using yet. One of the technical capabilities of the PPM is that Internet radio streams (whether Internet-only or exact or inexact AM/FM simulcasts) can be encoded to be picked up by the meters.

This means, at least theoretically, that any popular webcaster could encode its streams and be measured in any PPM market (although the smaller number of respondents used in each market might require a multi-city sample for any kind of reliable results).

In any case, the “less than 1%” figure was clearly wrong, as Arbitron readily acknowledged yesterday that it ignores Internet-only properties and exact Internet simulcasts of AM/FM stations. (Mark Ramsey also explains this well in his Hear 2.0 blog, here.)

share:  del.icio.us.  post this at del.icio.us  Reddit  post this at Reddit  Digg  post this at Digg  Yahoo   post this at Yahoo! my web  Wink   post this at Wink  Windows   post this at Windows Live  Google  post this at Google Bookmarks  Newsvine  post this at Newsvine


Commenting is closed for this article.

AccuRadio is powered by...
Conference schedules
Sep. 12 SF Music Tech Summit: San Francisco, CA
Sep. 13 RAIN Summit Chicago @ NAB/RAB RadioShow: Chicago, IL
Sep. 14-16 NAB/RAB RadioShow: Chicago, IL
Sep. 24 IBS Radio/Webcast Conf.: Chicago, IL
Oct. 6-7 Digital Music Forum West: Los Angeles, CA
Oct. 13-14 Digital Content Summit/Music: New York, NY
Oct. 18-22 CMJ Music Marathon: New York, NY
Oct. 27-30 CBI Fall Convention: Orlando, FL
Nov. 5 IBS Radio/Webcast Conf.: Boston, MA
Nov. 19 IBS Radio/Webcast Conf.: New York, NY
Dec. 3 IBS Radio/Webcast Conf.: Los Angeles, CA