A Starving Crowd !
A Starving Crowd !
A Starving Crowd !
You are about to read eight pages of information that can save you
and/or your direct-marketing clients from financial disaster.
And anyway, it's all a "setup" to pave the way for me to explain a
financial "life or death" concept that will be revealed near the end of
this letter.
Onward. No more messing around. Let's dive right in. Listen: As you
may or may not know, every once in a while I give a class on copywriting
and/or selling by mail. During these classes, one of the questions I like
to ask my students is: "If you and I both owned a hamburger stand and we
were in a contest to see who could sell the most hamburgers, what
advantages would you most like to have on your side to help you win?"
The answers vary. Some of the students say they would like to have
the advantage of having superior meat from which to make their burgers.
Others say they want sesame seed buns. Others mention location. Someone
usually wants to be able to offer the lowest prices.
And so on.
"Is...
A Starving Crowd!"
And so on.
Onward.
Comments: This is a little better, but not nearly good enough. High
income areas are, incidentally, easy to identify because several
companies have compiled statistics on every zip code in the United States
and they can tell you, with great accuracy, the average income per person
in each zip code. They (these companies) can also, by the way, tell you
the average education level, average age, how much is spent (per capita)
on automobiles and a bunch of other stuff.
However, as I said, this still isn't nearly good enough. For one
thing, not everybody who lives in a high-income area has a high income.
Some of these people might be live-in maids or gardeners or some other
type of domestic servant. (Come to think of it, with the prices they
charge, California gardeners at least ought to be wealthy!) Some of these
people may have money but are not interested in investing. Some of them
may buy investment books from a bookstore but never by mail. Some of them
may have money they are inclined to invest but will only invest in areas
other than the stock market where they already have expertise.
And so on.
Comments: Not bad. We are now getting into areas where we at least
have an outside chance. Whether they are interested in the stock market
or not, we can't know, but, in any case, if they are interested, they
probably have the financial ability to do something about it. There's no
doubt about it. This group of people is certainly more likely to respond
to our pitch than the first two groups but, as you should soon see, we
can do a hell of a lot better.
Comments: Bingo! Now we're cooking. These are upper income people
who have purchased by mail a product similar to ours. What could be
better? This is just about as "hot" of a list as we can get! Or is it?
Actually, it is not. Let's keep trying.
Almost.
Possibility #10: There is, however, one group of people that will
respond far better than any of the other nine I have described. What is
it? I'm sure you already know.
Comments: All other things being equal, your own customers should
respond stratostrophically (is that a word?) better than any other list
you can get. Of course, there is one caveat: THEY MUST BE SATISFIED
CUSTOMERS!
So there.
Let us press on. What you have just read was designed to give you
(and me) a "refresher course" on those factors that make some lists more
responsive than others. With those factors in mind, we can safely come to
the conclusion that there are three main guidelines we can rely on when
we are picking lists to test. These three guidelines are recency,
frequency, and unit-of-sale. A brief explanation follows:
End of kindergarten.
Now let's get down to it. All you have read so far is just preamble
to what I really want to teach you in this month's letter which is the
real way to test a mailing list and...
No problem. The list broker contacts the list owner, gives him the
order and, quick as a flash, our little DM hero has his 5,000 names to
test and he does so using his trusty control package.
And guess what? The results are spectacular! Greed glands begin to
secrete and a continuation mailing is scheduled.
Now look, our DM guy is no dummy. He knows better than to order all
the rest of the 500,000 names without making another test. What he does
instead is he orders 50,000 more of the names and he tells his broker to
make sure these names are also a random ninth name sample. He also tells
his broker to make sure the list owner keeps a record of which names he
is mailing so those 50,000 can be deleted if he decides later to mail the
entire file.
So far, so good. He gets his 50,000 names, puts his letters in the
mail and sits back to wait for the results. And those results, when they
come in, are not bad. Not so good maybe as the spectacular results he got
from the initial 5,000, but still, quite respectable.
And so, he orders all the rest of the names (445,000 of them) and he
spends something like $160,000.00 to mail his DM letters to them.
And wham! It's like a sucker punch to the gut! His results are
terrible. His business nose-dives right into the toilet. He calls his
list broker and he calls the list owner and screams bloody murder. "How
could this happen?" he says. "You must have cheated me! You gave me one
list when I tested and another when I rolled out!"
"Not so," says the owner. "You must have changed something in your
mailing. Or maybe the time-factor made the difference. After all, it was
60-days from your first test to your rollout and, as you know, time
changes everything."
And what about that 50,000 name continuation? What did our guy get
then? Quite simple: he got the rest of the best of the names on that
list.
And what did he get when he rolled out the remaining 445,000 names?
Well, it was the same list. Only those were the older (less recent)
buyers, the ordinary buyers on the list, the people who only made one
purchase instead of several, the people who bought the cheaper items the
list owner had to offer. It happens every day. Day in and day out. You
see, many people in direct marketing make more money renting their list
than they do selling their products and therefore they do everything they
can to maximize rentals.
You know, in a way, these sleazebags are like drug dealers; they
give you the pure stuff first in order to get you hooked and then, when
you go back for more, you end up snorting talcum powder.
Forgive me. That was a poor analogy. (It works for today though,
doesn't it?)
So anyway, the big question is, "What can we do to make sure this
never happens again? Is there any practical way we can protect
ourselves?"
Step 1: If you are testing a new package, go ahead and order a ninth
name random sample of 5,000 names just like our DM hero did in the above
illustration. Then, mail your mailing and keep track of the results.
Now, if your results are poor, forget the list. Trust me, you have
just mailed the best names that list owner has to offer. However, if your
results are good (or if you are using an already proven control package),
you then go to step 2.
Step 2: You now mail a continuation of 50,000 more pieces but, this
time you do not ask for a ninth name random sample. NO! This time you
choose three or four or five or six states and you tell the list owner
you want all of his customers in those states. Incidentally, as you
probably already know, there is usually a "state count" on the back of
the list card so, what you do is, you look at those state counts and pick
a few states that will yield a total number of names that is somewhere
near the 50,000 you need for a sensible continuation.
What will happen when you order names this way? Well, most likely,
your broker (out of ignorance) and the list owner (out of greed) will
warn you about getting skewed results because of a geographic bias.
And, they are right. You will get a geographic bias. But, most
likely it will be a "manageable" bias and it won't be:
Listen: I know that testing this way doesn't sound as mysterious and
complicated as some folks (folks who don't have to meet a payroll) would
like it to be. These people can talk merge-purge, regression analysis,
demographic overlays, and all the rest of it for as long as they want
but, I don't care: If you do it precisely like I just said, at least you
or your client are less likely to get crucified when you first roll out
to a new list.
O.K., now that that's out of my system, before I sign off, I'm
wondering, did any of you figure out why some names (as described in last
month's letter) worked so much better than others?
Don't feel bad if you didn't figure it out because I didn't either.
I got the solution from my ex-wife Nancy. It happened one day when I was
sitting with a mess of result figures from the individual surname
mailings when Nancy said,
Sincerely,
Gary C. Halbert
In-Print Salesman
P.S. Next month I'm going to reveal 20 ways to make your promotions work better
without changing a word of your sales copy!
P.P.S. Thanks for all the nice letters and phone calls I got on last month's
issue. Your praise went right to my head and made me even more
insufferably pompous, arrogant and conceited than ever!
Seriously though, thanks for the nice feedback. I really appreciate it!