MAILING LISTS

The single most important factor in selling by mail is the mailing list. According to a Dun & Bradstreet online report, other factors being equal, the list contributes 60 percent to the success of your mailing.

Offer is given 20 percent; copy, 15 percent; and format (design, envelope, art treatment, etc.), 5 percent. The art and science of selecting lists lie in our ability to match the recipient with the offer—to mail only to those most likely to buy. No matter what we are selling—no matter how appealing the offer—if the recipient is not in the market for our products, nothing else matters. We’re not going to sell dog food to cat fanciers; we’re not going to sell chain saws to apartment-dwelling couch potatoes. From the marketers’ standpoint, a mailing—as well as any other promotion—becomes junk when it is targeted at the wrong audience or when the quality of the mailing gives the wrong image of the sender. Note that the wrong image may be too rich as well as too poor! There is no better prospect than a satisfied customer. The most responsive mailing list is almost always made up of your present customers and clients, provided that such a list exists and that they have been well served in the past. That is why so many businesses use such ingenuity to gain your name and address when you pay by cash. You’re their best prospect for mail order, too.

When, however even your customer list seems to contain little useful information, some basic research is in order. People with like lifestyles tend to purchase alike. The key to selling is to determine which lifestyles match what you are offering. There are thousands of mailing lists available through list companies and brokers. (Chicago’s business-to-business Yellow Pages list about 100 such companies.) The more we know about the meaningful characteristics of our current customers or clients, the easier it will be to find likely prospects by matching their customer profiles to ours. What’s meaningful may not be a matter of common sense, so discuss with several mailing list professionals how they propose to help you find out and what they charge for their service. Then check the advice you get in a test mailing.

• Fewer than 5,000 names is not a statistically valid test. No matter how large the list, a 10,000-name test is almost always adequate.

• For consumer lists, services exist that can match addresses with socioeconomic census data. Use the nth-name system of every 20th, 50th, or 500th name, depending on the size of your list, for a test mailing. You’ll get a wealth of information on these people’s lifestyles.

• For business and professional lists, existing directories can tell you everything you need to know—from an individual’s specialization, age, income, family, and automobiles owned to a corporation’s history, profits, officers, and credit rating. If a certain piece of information seems pertinent, it’s there for the matching.

• A mailing list specialist, preferably experienced with your type of business, can help you analyze your lists. Before you request such help, get firm, written cost quotations and current references. Check them out!

For New Businesses

When your business is new, without customers or clients, there are four possibilities regarding mailing lists: 1. Lists exist to fill your needs. Your market is so well defined that available lists are all you need to get started. 2. Your market is hidden within a larger audience for which there are lists. 3. Lists exist, but they are not available to you. 4. No known list exists that will fill your needs. Let’s consider each of these in more detail. Lists That Fit With lists that fit, your most important decision may seem to be whether to use the list as a whole or to start in with a test. But before you do either of these, check on the percentage of previous mail order buyers in that list, no matter what they have bought or what you are selling.2 Have the sources of the list give you those proven direct mail buyers only. (Forced subscriptions to association magazines and newsletters don’t count.) It’s almost always easier to sell such buyers a second time than to sell nonbuyers the first. If no list of previous buyers exists—and you feel that you must sell by mail—test! Lists That Hide Let’s assume that you have a special racket for overweight, left-handed tennis players, but no mailing list exists for such individuals. A broader “umbrella” list of all tennis players, however, is available. Should you try to sell your submarket within it?

The answer involves the same analysis you should do before any other mailing. Do as little blind guessing as possible. Do a market breakdown; that is, work with what you know or can learn—for instance, the percentage of left-handers in the general population and the degree of overweight seen at local tennis courts and clubs. Although this type of “knowing” is far from certain, it’s better than sheer guesswork, and it lets you go to the next step: the calculation of testing costs and the application of test results as a predictor of complete mailings. A number of easy-to-use formulas exist for this purpose (see pages 105–107). Lists That Remain Private Many mailing lists are so valuable to their owners that they are never made available to anyone else. This is especially true of business customers and prospects, such as Collectibles for whom their list is their most valuable promotion property. General consumer lists tend to be less jealously guarded, for two reasons:

1. Increased use of a mailing list tends to build a larger universe of frequent buyers. As different kinds of products and services are offered, more recipients get into the habit of ordering by mail, making the list increasingly valuable as a source of proven mail-order buyers.

2. For many owners of a list, the income from renting the list is a major factor in their profitability. Suppose a list of 200,000 names generates net rental revenues of $25 per thousand 15 times a year. How much of the product would have to be sold to produce the same number of dollars in net income? The fact that some lists are not generally available tells you that people have been able to build them for themselves. If you have a list of your own, perhaps you can trade, rather than rent. If not, perhaps you can build your own list, too. When No List Exists Even when no list exists for a particular market, it’s probable that the names are there if only you can find a way to get at them. It is possible that no one else has previously wanted just those names badly enough to create a list. But it’s more likely that gathering the list would have been too difficult or costly. When no list exists, test using other media, including the Internet, to generate leads or sell. When no list exists, think very hard about the practicality of making direct mail the key to your selling effort.

Things to Know about Using Lists Eliminate Duplicate Mailings Where Practical When using more than one list, the possibility of duplicate mailings becomes increasingly likely. Through a computerized system called merge/purge, most of this duplication can be eliminated. But before deciding to use this program, discuss the process and costs with both a mailing list expert and a mailing service. Cost alone— lists, printing, mail handling, postage—may not be the key factor: Recipients may become so annoyed at receiving multiple copies of the same thing that they will consider it junk mail and pay no attention to it. Testing here, as elsewhere, will be the best way to discover what, if any, increase in response is achieved—at what cost—by merge/purge.

Much more about this is in Article 16. Some Legal Limitations on the Use of Mailing Lists When using a list other than your own, the rental agreement almost always calls for one-time mailing use only. You may not copy the list or any part of it, or use it for any other purpose, unless agreed to in writing by the owner of the list. After testing, consider negotiating for multiple use of the most successful lists at reduced cost. Do not, however, pay for such multiple use in advance. First make sure the full mailing lives up to the promise of the test return. Rental lists are “seeded,” that is, they include a few names specifically added to discover unauthorized use. Any response to your mailing, however, whether it is an order, an inquiry, the acceptance of a premium, or anything else, makes the respondent’s name yours to use in the future. With many mailings, you justify their cost by gaining repeat customers, rather than one-time selling of a product or service.

Customer or One-Time Buyer Know what you need to gain from mailing—one-time sales or long-term customers—and evaluate the results accordingly. But beware of paying for customers and then giving them a product or service that reduces them to one-time buyers, or budgeting for customers while having nothing more to sell. Direct response may be the one way of doing business where you can sometimes lose a little bit on each sale and then make it up in volume—but only if you also know how to play that tune. Know what business you are in and your capabilities!

THE ABSOLUTE NECESSITY FOR TESTING In direct mail, or any other kind of direct response advertising, the likelihood of getting things just right and most cost-effective on the very first try is quite small. We can, of course, and often do produce profitable mailings on just one try. But most cost-effective is most often the result of testing—a process that never stops in many direct mail organizations. In testing, the version (“package”) that does best is called the control. Everything thereafter is controlled by (evaluated against) this package. Most of us tend to use “control” synonymously with “success.” Its actual meaning is “the best results thus far,” which are subject to change with every mailing.

Test One Thing Only Changing one or more factors in your mailings may give you a better response, but it may also have no effect or decrease returns. Even an increased response may have to be balanced against higher mailing costs, and lower costs may have to be balanced against fewer sales. It is quite common to test totally different mailing packages against each other—one or more quite expensive and the others somewhat plainer and less costly. But when testing to see whether a specific package can be made more effective, change only one thing if you wish to understand the results.

• Changing the offer.

• Changing the price.

• Changing to a less costly two-color mailing package. Because it takes a mailing of 5,0004 pieces to generate fairly predictable results, you will need 12 different mailings totaling 60,000 pieces for your test. The reason we need so many, is that we are testing one specific factor in each mailing, yet testing that one factor against all the other 11:

• Each of the prices is tested against all combinations of offers and colors.

• Each of the color arrangements is tested against all combinations of offers and prices.

• Each offer is tested against all combinations of colors and prices. Should one of the offers be tested at random—for instance, the four-color offer A at $15.95 against the two-color offer B at $22.50—one will probably sell better than the other. But you have no way of knowing why—whether it was the offer or the price or the color that was responsible for the increased sales. Nor will you know which factor(s) you might change for even better results. You seem to need to perform all 12 tests to get your answer! But don’t give up on testing as being too complicated and costly. Help is on the way immediately after the next paragraph.

When to Stop Testing As you see, testing one more price will add 20,000 units for four more packages. Testing envelopes against self-mailers, or first-class postage against third-class postage, however, will double the number of units needed—from 60,000 to 120,000. You’ll run out of names or money or both before you run out of tests, so let’s take another look at testing.

Five Ways to Approach Testing 1.

Test everything. Test your product or service in exactly the way that was described in the last few pages. It’s what the largest, most successful mailers do. 2.

Decide what might really make a difference, and test just that. For instance, of the 12 options, test only the four-color offers A and B at $15.95. If one is successful, or close to successful, test the other two prices for the same package. Now test what has become your control against that same package in two colors. The result is that you’ve mailed 5 sets rather than 12. In other words, you’ve mailed 35,000 fewer pieces and saved at least $6,000 just in postage ($3,000 at nonprofit rates). For mailings where four colors might be perceived as an extravagance—to nonprofit organizations, charities, hospitals, and so on—test two colors or one color first. But do try four colors also. Don’t assume; test. 3. Test fewer than 5,000 names. When you really need to test how a large number of changes will influence results, and restrictions on budget, time, or names make testing 5,000 names impossible, test 2,000 names instead. Then take the top two or three results, and do a “real” test of 5,000 each. If you can possibly avoid it, don’t go from a 2,000-name test to a “rollout”—a mailing of a much larger portion of your list or a complete mailing of the entire list. If you do not have time for a real test and results are important, either don’t mail, or use a professional—and pray! 4.

Use telemarketing to test before you mail. If your primary concern is your offer or price, give serious consideration to a telemarketing premailing test. Properly structured, it’s fast, accurate, and the way to find out why someone doesn’t buy, as well as why they do. (See Article 9 for a more detailed discussion of telemarketing for premail testing.) 5.

Use print media to test before you mail. Find print media—magazines, newspapers, and newsletters—that go to the audience you want to reach. Check with each one to learn whether it accepts preprinted advertisements on postcardweight stock. You want the best results from your test, and a reply card almost always increases response. That card, in combination with a toll-free number, should give you the best results. If card stock is not acceptable but regular paper inserts are, consider an order form similar to what is used in most catalogs, preaddressed and, if publication policy permits, postage free.

How to Prepare Your Print Media Test Do the following:

• Use a freestanding insert for your test.

• Check with each publication on the size, format, and delivery date for your insert. Newspapers will probably accept 81⁄ 2_ _ 11_. Magazines not only vary in page size but usually require extra paper for trim. If unsure how to handle all this, get help from a printer experienced in this field or a print media production pro.

• Prepare each test as an insert with all of the four-color or all of the twocolor offers printed together. If you are printing on postcard-weight stock, for easy removal, have the card edges perforated.

• Use a department number to code each reply form. Use the same department codes for telephone response and add “M” for mail and “T” for telephone. Keep the system simple. If you’re not staffed to handle telephone orders, see telemarketing services in Article 9.

• After printing, have the 12 versions separated by your printer or bindery and collated sequentially, that is, into the order 1, 2, 3, 4 . . . 12. This gives you random, valid testing. What you must not permit is using all the #1’s first, then all the #2’s, and so on.

Testing Just Two Variables If only two variables are to be tested, discuss, with each publication, the practicality and cost of running an A/B split advertisement with a reply card. (See page 35). This format, especially when also used with a toll-free number, gives statistically valid results. If you do not have a toll-free number, check with a telemarketing company that offers this service. They’re in the Yellow Pages. A word of warning: Many professional publications build their circulation by sending multiple copies to the same destination, although not often to the same person. Persons within the same organization will thus get different versions of your offer! If they compare and contact you about the differences, tell them you are running a test. They’ll understand.