Yesterday afternoon Arbitron issued a statement advising against comparing its own survey-based PPM audience estimates to those of "Internet music services" based on server logs.
In its statement, titled Thoughts on Comparing Audience Estimates (it's a .pdf file, here), Arbitron (1) insists that audience estimates of broadcast listening (many people listening to the same thing at the same time) can't reliably be put side-by-side with estimates of webcast listening, where each listener gets his own stream.
Arbitron also advises (2) against using webcast audience estimates that don't include explicitly-cited limitations and "detailed description of methodology" for making their estimates. Next, (3) with "many Internet music channels... there appears to be no way of confirming if anyone is on the other end throughout the session." And, (4) those Internet services can't always verify whether "self-reported registration data are reliable and that users do not have multiple accounts."
Read a more point-by-point summary of the Arbitron paper from Tom Taylor in Radio-Info here.
Consultant Mark Ramsey defends the veracity of server-side measurement and takes apart Arbitron's arguments here.
"Perhaps we should be bringing Arbitron up to date," he wrote, "rather than blowing dust onto metrics which are based on every user with 100% accuracy, not a smattering of sampled users with sketchy accuracy."
RAIN Analysis: The timing of this statement from Arbitron is pretty awkward, it would seem. Arbitron needs to walk a pretty fine line here. On the one hand, it's attempting to discredit audience estimates based on the data culled from servers about "when" and "how much" those servers send out into the ether. Yet, it needs to quell a brewing client rebellion among broadcasters over its own Total Audience Measurement service -- a service that will reportedly estimate listening based partially on server log data. Stay tuned. -- PM