"John Kasupski" wrote in message
...
On Sat, 7 Jun 2008 14:35:40 -0700, "David Eduardo"
wrote:

It's really very simple. Ask anyone with access to ratings data to do a
run
on 12+ cume share for a combo created out of the three mentioned AMs in
SF.
They reach about 1 in ten persons, no more.

Oh, come on now! I've pretty much stayed out of this so far but the
broadcasting industry has far less of a clue with respect to the
demographics and numbers of its listeners than it and you would have
us believe - and this applies to television as well as radio.

Both industries use sampling techniques used in every other facet of
American business to measure consumer response. The methods and sample sizes
are perfectly adequate for the intended purpose.

To begin with, ratings are based on paper surveys, which of course are
kept by only a small percentile of the total number of listeners in
any given area, who are participating in the ratings "sweep"

There are no sweeps in radio in about the top 100 markets... measurement is
constant. In the new electronic PPM, still being perfected, measurement is
instantaneous.

Again, the samples are adequate for the purpose. Replication testing shows
little if any gain if the sample is increased, in fact.

(Arbitron
typically passes out between one and four thousand paper surveys in a
given market) -

Actually, they are weekly diaries and there are between 500 (market 298) and
7500 diaries per survey (4 times a year in most pf the top 100 markets)

and then, of course, the results are tabulated from
the surveys that listeners return (How many listeners simply toss them
into the nearest waste basket as soon as they receive them?).

Some do, but Arbitron anticipates this. They know the non-return rate and
recruit enough extra diary households to compensate for the unreturned ones.

Whast this means is that you are getting data from only a fragment of
a fragment of the total potential audience.

Wrong. People are recruited in excess, so there is no "fragment of a
fragment." If they need 7500 diaries back for New York City MSA, they may
send out 10,000. They know so well who returns and who doesn't that they may
send out 50% more diaries to 18-24 year old men, but only 5% extra for 45-54
year old non-Hispanic white women. After all, Arbitron has been doing this
for 43 years.

And a "fragment" is called a sample. Just as they don't drain all your blood
for a blood test, they take a sample of a cross section that is totally
proportional on age, sex, ethnicity, geography of the market, etc., and
project it into the universe of listeners. This is the same sort of thing
Proctor & Gamble does when developing new products or finding ot why a
competitor is gaining market share.

This may fool broadcasters
(who could really care less what the listeners want and are only
interested in selling advertising),

We can only sell advertising if listeners are interested in our stations and
listen a lot. Radio is sold "by the listener" so those listeners are awfully
important.

and it may fool advertisers (who
could really care less what the listeners want and are only interested
in how many listeners their ads will reach), but it doesn't fool
listeners - many of whom change the station the instant the
commercials come on anyway,

Actually, listeners are much more likely to tune out due to a bad song than
commercials. Listeners know radio is ad supported, so most put up with the
ads but not with bad programming.

so when a survey asks them if they heard
the Burger King commercial on WWTF at 8:45 PM on Saturday night, the
answer is no, not because they weren't listening to WWTF at 8:44, but
because they STOPPED listening to WWTF at 8:45 when the commercials
came on.

No survey asks that anyway. Nice try.

Of course, the surveys also rely on the listeners remembering
everything they listened to during the period.

Listeners only have to remember what station (by calls, frequency or program
or DJ name) they listened to, not the details. And the new PPM does not need
anyone to remeber anything.

This from people who
generally have no idea who the Chief Justice of the U.S. Supreme Court
is and can't remember what they had for lunch yesterday.

If you listen to a station many hours a week (8 to 10 hours being typical
for one's favorite station) it is likely you remember their frequecy at
least. Otherwise, how would you find them.

And don't bother to tell me about the new PPMs, either. It's already
known that they have problems measuring stations with niche audiences,
and their sample size is even smaller than that of survey-based sweeps
(and anybody who knows Jack Schmidt about statistics can tell you that
a good way to make bad decisions is to base them on numbers that are
the result of too small a sample size). Also, like the paper surveys,
these devices measure exposure, not attention.

Advertisers only want a measure of exposure. And the PPM measures niche
audiences fine. In Houston, the only accredited market with the PPM (there
are only two total markets running, so you are generalizing on scant data)
the top 2 stations in February and March were niche: Majic and The Box.

Here's how a typical commercial broadcast radio listener behaves
today: Turns on the radio. Whatever station the radio happens to be
tuned to when it is powered up is what the listener hears first. If
the listener is looking for a particular program (maybe the broadcast
of that day's baseball game), and it's on that station at that time,
fine, otherwise ZAP the station gets changed.

70% of listening is in the home or office. Most people are not close to the
radio or attentive enough to it to change station each time commercials come
on. Only in the car, with the convenience of buttons, is there station
hopping, but it is nowhere near as extreme as you would immagine. People
tend to have a coiuple of favorite stations, and do not do much other
seeking or switching.

Let's say the listener tunes into...Rush Limbaugh for example. At the
top of the hour when they take time out for the commercials, guess
what? ZAP the station gets changed, listeners know EXACTLY how long it
will be before Rush comes back on, and they don't bother listening to
the crap that's on in between.

This hardly ever happens. It's not seen in minute to minute PPM results nor
in diary based results over the last 18 or 19 years.

If the listener wants to listen to rock music and the station's
playing rap instead, ZAP the station gets changed, and keeps getting
changed until the listener finds music that's acceptable to him/her.

That would only happen if you moved to a new city. Listeners know what
stations they like and what ones have the music or features they prefer.
They seldom listen to any other, unless encouraged by advertising or
recommendations. In fact, the average person only has between 2 aqnd 3
stations they listen to for more than an hour a week... they just don't go
wandering around looking for other things once they have decided on the few
they like.

If the station's playing rock, and the listener wants to hear rock,
the listener stays...until the first commercial or a rap song comes on
and then ZAP the station gets changed.

I don't know of any stations that play rock and rap; those are separate
formats.

That's the problerm with your ratings - you have no numbers that
matter. As Thom Mocarsky, the vice president of communications at
Arbitron, stated in Media Life Magazine, "Neither the diary nor the
PPM measures attentiveness."

They are not supposed to. ABC does not measure readership of a paper,
either. All advertisers expect is an indication of how many chances they get
to make an impression.