In addition to list owners and vendors, who rent their internal opt-in lists for
one-time use, we also can add another source to find opt-in lists the e-mail list
broker. A broker is not beholden to one list owner or vendor. Instead, it has access
to dozens of vendors and, within them, hundreds of lists. It also can help in the list
planning and research process, and can negotiate pricing. A broker earns its keep
by charging a commission; yet more times than not, you do not pay any more than
if you had gone directly to the list vendor because of the pricing method that many
list vendors use. Most list vendors charge a gross price, whether they’re dealing
with an agency, a broker, or directly with the advertiser. The gross price is the cost
of the list marked up by a 10 to 20 percent commission.
Oftentimes, a broker can
negotiate that price down, while still earning its normal commission. Therefore,
the benefits to using a broker are twofold: First, you have one-stop access to hundreds
of different e-mail lists, and second, you can also, more times than not, get
a better deal than if you negotiated directly with list owners.
We’ll delve into some nitty-gritty details of opt-in e-mail lists and how to
find the ones best targeted for your promotion; but before we do, let’s define a few
direct marketing terms that have transferred to e-mail and are specific to lists and
list research.
Universe. When you first begin the search, your job is to find as many
sources as possible for your target audience. If you have a broad audience, you
may be able to reach a potential audience of 500,000 or more people across various
e-mail lists. Conversely, if you have a small audience, you may only be able
to reach 50,000. That number is your universe, or your gross potential audience.
A universe also can refer to the total quantity of a list.
List test. New lists that have never been used by your company always
should be pretested in small quantities. Even if you determine that a new list has
50,000 quantity and your budget can withstand the costs, the general rule of thumb
is to test a small quantity of the list first. If it pulls a strong response, you can then
go back to it at a later time. A minimum quantity of 5,000 is recommended for a
list of 50,000. Generally speaking, the test quantity should be at least 10 percent
of the segment or list that you’re testing. And to be statistically valid, you should
aim for at least 50 responses.
Continuation and balance. After a test mailing has gone out, and certain
lists have been deemed worthy by their response to be e-mailed again, a marketer
can choose to “continue” on that list if he doesn’t want to rollout to its entire universe.
Example: If a list has a universe of 50,000 and in the first e-mailing, the
marketer tests 5,000 e-mail addresses within that list and it does well, the marketer
can e-mail the balance of 45,000 in the next mailing. Or, if the budget is smaller,
the marketer can take a continuation of that list a fresh set of 5,000 addresses in
which the initial 5,000 is omitted.
Rollout. This comprises all the e-mail addressees within a campaign that
are sent at one time, including list tests, message, creative and offer tests, and any
other variables.
Statistical significance. In order to get a solid reading on a list, you need
to make sure you have enough names so that any response you get can be applied
to other promotions. For instance, if you glean a 5 percent click-through rate on a
list test of 5,000 (where the total universe of the list is approximately 20,000), you
can pretty much count on the fact that new names within that list i.e., names that
have not yet e-mailed will give you similar results. However, if you get similar results from a list with only 500 names, you won’t be able to e-mail that list again
with complete confidence that it will yield similar results because the quantity was
not high enough to make it statistically significant.
Nth. This term refers to a specific randomly selected quantity, normally
used for test quantities of a list. For instance, a list may have more than 50,000
names/e-mail addresses on its file. For testing purposes however, you may choose
to pull off an nth of 5,000. If done right, it will not be the first 5,000 names on a
list and it won’t be the last 5,000 it will be names randomly selected across the
entire file. This is to ensure that the test quantity represents an accurate sampling
of people within the list.
Now that we’ve determined the basics, let’s explore some available categories
within lists. Again, the Resource section at the back of the book lists sources
of lists, including vendors and brokers, and also includes a brief description of
what categories of lists they specialize in, if any.
|