My name is Jordan Mendell, and I'm addicted to charts.
I just can't get enough of them.
The amount of pie consumed.
The number of things Arnold Shwarzeneggar has been trying to get to.
Chances of an important call being returned
There are any number of potential interesting data points available in Internet radio. While you have no interest in charting the number of times Arnie yells "Get to da choppa", you may have a need to chart the differences between listening time of people on Rock stations versus Classical stations.
I believe one of the biggest value propositions between Internet radio and broadcast radio is that the former is a 100% measureable medium. Broadcast radio is a unidirectional medium, where the tower simply broadcasts the audio into the air and the actual number of people who hear it is completely unknown (or estimated). Internet radio is different because there is a two-way communication going on between the listener and the server sending the audio streams. Therefore, you know *exactly* how many people are listening to that stream.
Obviously I'm biased because I'm part of a company who's business model is based on measuring Internet radio, but why would you use estimates when you know exactly how many people listen?
Some people say that there is a difference between "knowing how many computers are connected and how many people are actually listening". This is the typical argument that there is a difference between counting listeners and counting computers. There is assumed to be a tiny percentage of people who repurpose Internet radio in commercial establishments such as restaurants, grocery stores and other places where there may be more than one person listening. This would indeed undercount those people. As it stands now, I have yet to go into my local grocery store and hear a local radio station's Internet stream playing.
I'm not going to bash any companies by name, but there is one who's name starts with an A has BITRO in the middle and an N at the end who states that putting 50 people meters in a large city will somehow tell you how many people are actually listening to Internet radio stations in that city. How absurd.
I was recently asked to give my opinion on a single station who's data was collected via PPM and Ando simultaneously over the course of a week in September. PPM reported that there were over 32,000 people listening to the station in question. Ando showed that there were an actual total of 1,746 people listening.
Let's make this absolutely clear.
PPM: 32,000 listeners
Ando: 1,746 *actual* listeners
I've been tracking listener data on Internet radio stations for over a decade, and to burst someone's bubble, there are NO STATIONS IN EXISTENCE who reach more then 32,000 listeners on a single channel station. Don't believe me? Hit http://rs6.net/tn.jsp?t=pjyxftcab.0.0.xyb5ilcab.0&ts=S0373&p=http%3A%2F%2Fwww.shoutcast.com%2F&id=preview between 12 and 3pm EST on a weekday. The number one station listed there (.977 The Hitz Channel) is the largest single channel Internet radio station on the globe and they peak on average at about 25,000 listeners.
The fact that radio stations are using this seriously flawed data to sell against and promising advertisers a certain reach based on these numbers is quite scary. Especially when the advertiser expects some level of response (clicks, calls etc). In other words, if a radio station sells to an advertiser who is expecting to reach (for example) 32,000 people, but in reality they are *actually* reaching 1,746 people, they will immediately see that their campaign is not performing, and that radio station may ultimately lose them as a client.
Ok, time for me to go. I've got to print new cover sheets for my TPS reports.
Till next week!
Jordan Mendell
No comments:
Post a Comment