|
7. Why Was Buried-Offer Testing Abandoned?
If split-run testing of two or more different ads with
the same buried direct-response offer is so effective, why has it fallen
into such disuse today?
The answer is shrouded in the mist of advertising history.
But we can hazard a guess: First, the development of radio, and then of
television seemed to require - and did produce - new, more elaborate forms
of copy research which came to be accepted as the state of the art.
Second, advocates of other methods of research bolstered
their arguments against direct-response testing with that we believe to
be a statistically fallacious argument. This argument was summarized in
a scholarly book published in 1936, Four Million Inquiries from Magazine
Advertising, by Harold Rudolph:
It might be argued that replies do not constitute
a representative sample of the magazine-reading population since only
a small percentage of any magazine's circulation is made up of potential
coupon clippers. For example. the average advertisement draws coupons
from less than 1/10 of one percent of the readers to whom it is exposed.
Therefore, in order to show an increase of 50 percent in replies, an
advertisement need only secure responses from an additional 1/20 of
one percent of the circulation. Is this a significant margin? In other
words, should one advertisement he considered superior to another because
it has elicited response from 1/20 of one percent more readers?
Our answer is that a 50 percent increase in response
cannot be a random result explained purely by the laws of chance.
In other words, if two identical response ads are given equal exposure
to two equal audiences over and over again, the total replies to each
would be roughly equal (with the plus or minus variations you would get
from flipping a coin).
So if the responses from ad A exceed the responses from
ad B by a statistically greater margin than you would get from coin flipping,
ad A is statistically superior. And these are responses from real prospects
in a real advertising environment, not artificially stimulated opinions
and responses from human guinea pigs who may not be willing to buy your
product no matter how good your advertising is.
Like all the other research methods, split-run testing of
ads with an identical buried offer cannot relate each advertising exposure
directly to sales. But it can measure and compare public reaction to the
advertising message at the precise moment when it is reaching the target
audience in the real media environment.
By testing A against B and the winner against C and the
winner of that round against D and so on, split-run testing can build
incremental improvement in the proven impact and believability of your
advertising.
Thus as the glaring limitations of other kinds of copy research
become increasingly evident, one of the oldest forms of advertising research
may become one of the newest-especially since it fits in so well with
the other requirements of MaxiMarketing.
|
