D Peter Maus wrote:

Eric F. Richards wrote:
[...]

Actually, what I've gotten from this discussion is that even if the
tracking methods -- 800 numbers, "Mention you heard it here on...,"
etc. have all shown that there are listeners all beyond the target
areas. What happens is that the sales department doesn't like data
that doesn't fit their assumptions, and dismisses it out of hand.


Actually, it's a lot less sinister than that. It's that there are not
sufficient numbers of them to be saleable to advertisers. The fact is
that few people actually listen to any given station out of the local
coverage area. Skywave listening is still going on, but not in saleable
numbers. There is no mechanism for selling a widely scattered irregular,
unmeasured audience. For an audience to be saleable, it needs to be
measured, and fall in to the correct demographic, psychographic, and
geographic areas. A zip code with less than 100 listeners, is
statistically zero. A zip code with an unreliable signal is of no value.
Believe me, if the numbers supported it, WLS would have a sales
office in Shreveport, Louisiana. But the only one regularly listening
to WLS in Shreveport, was me, in 1984. There were a half dozen of my
friend in St Louis, who listened to WLS. Most of them were in Radio.
Most would prefer to listen to KXOK. The signal was stronger, clearer,
and more reliable. Even in the 60's there only pockets of listeners to
skywave activity. Widely scattered, occasional listeners are of no
statistical presence. And not saleable.


But that still misses the point. The idea is not to target
Shreveport, but to sell to a company that can target Shreveport,
Montgomery, Pensacola, Ft. Walton Beach, Huntsville, Birmingham, and,
yes, those of us like me in East Overshoe that just *might* be looking
for a reliable mail-order company in a large city that has the SLR I'm
looking for, because there are no local places to go.

Again, the idea isn't to sell pizza from Chicago to Huntsville. Or,
to sell Huntsville pizza on a Chicago radio station. The idea is that
there are advertisers who appeal to any location in a large geographic
area, such as J&R -- or any other business that sells mail-order --
and put their ads on the air.

I wouldn't be surprised if the 100 here and the 50 there over ALL of
the skywave coverage area added up to between 5 and 10% of your
listener area. Would *you* want to tell the PD that the latest
Arbitrons showed a 5% drop in listenership in the evening?

But that is what you are doing -- assuming that the 5% is
statistically insignificant because you are looking at it in terms of
listener density per geographical unit.

....and its not like those listeners are costing you extra, in terms of
station expenses -- you aren't increasing power for the benefit of
those 5% -- you would, however, be selling to them an ad that is not
targeted to a geographic area.

The bottom line is that there's a bottom line. And anything that
can't materially affect it is not considered.

But your model is flawed. It assumes that a low geographic listener
density cannot be sold to, but that's only true for
greographic-sensitive products, like pizza and the local bar or Olive
Garden.

If that model was used on network television, there'd be no network TV
ads, but there are. And somehow, network TV muddles on.


That's the nature of Radio in the US. It's always about the money.


In this case, there's money to be made, even money being made, and it
is ignored. You said yourself, earlier, that if I call in from East
Overshoe and buy, say, that $1500 SLR, and say, "I heard it on WABC,"
they'll throw the data out rather than count it, because, *by itself*,
it's statistically insignificant.

....except that there are 100 sales thrown out that way, and they are
as an aggregate, statistically significant. All the East Overshoes
don't get their own pie slice, but put into the "Other -- Skywave"
slice, they are pretty big.



People are willing to do business cross country. And advertisers buy
national radio. But radio is SOLD according to local numbers.

And that is where the model is flawed.

Skywave
numbers are not statistically present, nor practically operable.
Literally, to few, to far between to be useful.

....and I'm saying they are, but you'll never know selling pizza. But
if you sell to a company that can take advantage of those distant
areas, they will.

How many Hallicrafters radios would have been sold if they only
advertised in Chicago newspapers?


Now, say you have listeners 400 miles away, well out of the
groundwave, and well into skywave. How many do you expect there to be in
any give zip code?

You don't. But if the local survivalist is the one with the program,
you sell the emergency preparedness web sites, Honda generators, and
enough colloidal silver to turn 'em all blue. Survivalists don't like
the city much, anyway.

Substitute "local survivalist" with "Art Bell" if you need a more
believable scenereo.

10? 100? If the conversion ratio of sales to
impressions is 1 in 10, that means to buy that station, one could expect
between 1 and 10 sales to result from a given period's advertising. 1 in
10 is very optimistic. So,the cost/benefit ratio is too high for that
buy. Now in the case of a mail order business such as, taking your
example, J&R, yes a clear channel station could produce a few sales here
and there though skywave listening, but consider, that the numbers,
again, are small compared to the local audience. And it's the size and
listening frequency of the local audience that sets the rate for the J&R
buy. Again, there is no statistical benefit to including the skywave
listener. Making any measurement of the skywave audience prohibitively
expensive.

Either way, they don't matter in the real world of Radio. Because
they produce no revenue enhancement.


....that your model will measure. If it doesn't fit the model it's
thrown out, because the model doesn't reflect it.

You aren't spending extra dollars to get the signal to East Overshoe,
it's just there. Now that it is there, put something out there that
will be available and useful to any and all the East Overshoes out
there, no matter where they happen to be. Leave the zipcodes out of
it, unless you start binning all the zipcodes from your skywave
listeners according to the demographics they represent. You have one,
here, two there, five here, one over thar, all of which happen to be
middle class males 18-45 and add up to 100 sales. Not saleable?


--
Eric F. Richards

"Nature abhors a vacuum tube." -- Myron Glass,
often attributed to J. R. Pierce, Bell Labs, c. 1940