There have been so many books and things about problem-solving that we figure most problems are already solved. So we need more problems to keep all those expert problem solvers busy.
Ken Rutkowski asked about how podcasts could work with webcasting, given that there are problems of bandwidth and measuring audience. Peter Yorke noted that the bandwidth problem can be solved with BitTorrent (at his site, DownloadRadio.org). Rob Greenlee says that the development of Podcasting is eroding listener streaming and notes that advertisers don’t know how to value Podcasting, because they don’t know whether anyone is listening.
In this episode, I start in simple mode. Radio and TV audiences have been measured for years, even though radio waves leave no tracks. The methods (surveys, polls, diaries) are well established and widely accepted. They provide important information that is not available from web tracking: Audience demographics. Ken and Rob may be able to ignore that issue because their content will select the demographics. But even there, it matters whether the audience listening to your stream is one or five and what their specific interests are.
I think web delivery is going to replace a lot of on-air distribution (and traditional cable, too). So I checked to see what Arbitron and Neilson are doing to stay in business. Both websites have web services right up front. Here are excerpts of a news release from Arbitron. (Dated 12-4-04).
an estimate of 4.1 million people a week, age 12 and older, listening to three major online radio networks.
rated the three charter subscribers – America Online’s AOL® Radio Network; Yahoo!®’s LAUNCHcast,; and Microsoft’s MSN Radio and WindowsMedia.com – during an average broadcast week in the month of October.
These results help us validate the size and value of online radio for advertisers on a level playing field with traditional radio,” said Andy Lipset, managing partner, Ronning Lipset Radio. “Online radio must use the same metrics used by terrestrial radio to be included in traditional media plans.
Cumulative persons (age over 12) per week. Yahoo, 1.9 M. AOL, 1.8M. MSN, 425K
The comScore Arbitron Online Radio Ratings service is based on a subset of approximately 200,000 U.S. participants within the comScore global consumer panel. Using proprietary and patent-pending technology, comScore passively and continuously captures the online behavior of these panelists, including online radio listening behavior.
Consumer panels seem to work well on the web. Here, comScore has apparently arranged to track users as they are streaming. I assume that they ask permission on sign-up and use the logon to know who is there. If they are sharp, they will send a query to a random sample (about 1000) of their panel from time to time. The query will ask how many people are listening to this stream and get a few demographics.
But this doesn’t work for podcasts (or for your car radio). Yahoo doesn’t support podcasts, and I suspect the others don’t either. Actually, Yahoo doesn’t offer any voice feeds on LaunchCast. It does offer voice feeds from NPR, but through its news pages, not through LaunchCast. I suspect some kind of change is coming. I don’t think Yahoo puts ads in its audio feeds, but it must be thinking about that. Otherwise, why would it spend money on Arbitron.
But let’s imagine that Yahoo decides to go modern and offer podcasts. (Such miracles do happen. Yahoo now offers RSS feeds.) Arbitron would merely adjust the query method. Send an e-mail asking for demographics of listener(s).
So where do we go from here? I think some aggregator will start working on a plan for measuring audience of podcasts. Not the content people because they would just duplicate each other’s efforts and will not be credible without audit. The aggregator’s plan might begin by contact with Arbitron, where they say, “This is an interesting idea and we are going to follow it. If Podcasting ever takes off, we will be right on top of it. All it takes is money
Moving on to Plan B, the aggregator could set up the initial survey. You want aggregation and automatic transfer to your device. Become a paying member ($15 per year) or join the panel to get free service and a chance to have your feedback change things. Joining the panel means you fill out an initial demographic survey and respond to occasional brief e-mail queries asking about your use of a recent download and about commercial products of interest to you. Maybe to aggregator attaches ads to the download (like the cable companies do). The aggregator could attach ads fitted to the member. (Is Google smart enough to see an opportunity there?
Would there be other commercial uses for data on what products some people want? Is there a patentable technology here?)In the long run, Arbitron would be needed because advertisers would want an independent audio of the stats. But if the aggregator has a good working system or patented methods to fit things to Podcasting, the aggregator will be able to offer the whole thing to Arbitron. With the comment, “All it takes is money.”