One of the most Frequently Asked Questions I get in regard to direct marketing is "What does it cost to do a mailing?" So let's take a look at the costs involved and other considerations such as testing, that affect costs.

Recently, a young man I know, whose smarts in the area of developing and marketing shareware I greatly respect and admire, maintained that one could do a direct mail test for about $10,000. I took issue, making a few of the points I'll get to here in a minute. He said, "But what if $10,000 is all someone has to fund a mailing?" I suggested that person take his or her ten Gs, spend a week in Vegas and see if he can quintuple it at the tables. If not, he'll at least have had a helluva week in Vegas, and he won't have blown his $10,000 on a likely bomb of a mailing. After all, what would you tell someone who wanted to enter the Grand Prix de Monaco, but all he could afford was a  VW Bug?

The point is there are certain minimum thresholds in direct mail that must be reached or all that will happen is you'll join the ranks of those who say, "We tried direct mail once. It didn't work."

The most you can mail for $10,000 today is about 10,000 pieces, assuming you did the creative yourself at no cost. That's for a typical, 5-part direct mail package: outer envelope, letter, brochure, order card, and business reply envelope (BRE).

So let's say you want to mail the 10K pieces. That means you'll be limited to mailing one list. Or at the most, two lists, because you need to test at least 5,000 names per list for any kind of statistical assurance. If your anticipated return is around 1% -- a safe bet for budgeting purposes most of the time -  you really should test 10,000 names per list.   

O.K., you have this exciting new widget and $10,000 and you're determined to test this great list someone sold you (rented for a one-time use, that is). You prepare your package and mail it in expectation of the infamous 2% response. But when you count the order cards, you only have 1%. Now you've spent $10,000 and what have you learned? That your Widget is a winner? No. That it's a dog? No. All you know for sure is that  wasn't the right list. Costly answer.

Of all the considerations -- and there are many -- that go into a mailing, list testing is far and away the most critical. If you can't find your market, it doesn't matter what your Widget is. And the only way to find your market is to test lists. The more lists you test -- at a minimum of 5,000 names each -- the better. With each additional list you test, you're buying yourself valuable insurance. Especially since only about 30% of the lists you test are likely to prove out. Now you see the odds you're playing with a single list mailing. You're better off in Vegas.

THERE ARE EXCEPTIONS.

Some years ago I did a package for a brand-specific computer magazine. The list was a "hotline" of people who had just recently purchased that particular brand of computer. Result? Thirteen percent response. It was an innovative package to be sure (6" square in silver mylar, looking like it contained a floppy disk). But the tight fit clearly didn't hurt.

Just this year I did a personalized, web press package for a local Bed and Breakfast. We mailed 10,000 names for $6000 and pulled a 3«% response. The list was a geographic selection (based on the owner's registrations) of Country Inns Magazine subscribers. These are upscale travelers with a clear preference for Inns and B&Bs.

At these prices, the package had its limitations, however. One color, for example (black on a tan stock). But because I'm a brilliant copywriter, the package worked, and the B&B started building a premo mailing list at better response rates than they expected.

A web press package is one that's printed on a single, continuous roll of paper, then cut and folded into a direct mail package. It offers the opportunity for personalization, usually in at least two places (letter and order form), but with the budget I had to work with in this case, I couldn't do color, so I chose a colored stock with black ink. A standard package costs considerably more.