There is one exception to continual testing. I would not test any direct mail during the two weeks prior to Christmas. The preoccupation with Christmas can cause vast fluctuations in D.M. response. For classified and display tests in December, adjust them the same as you would for other slack months.

Price testing is the very first testing procedure for a newly marketed item. You can usually arrive at a tentative price range by comparison with existing products or services. Once you have established this range you can set up a three price test.

For number one you should use the lowest reasonable price. Number two should fall approximately in the center of the range and number three should be predicated on the maximum of your range. This “3 position” test will quickly tell you where your most responsive price occurs. It may well NOT BE the lowest price tested.

When these results are in, you can begin computing your most advantageous price. Of course, total [number of sales doesn’t tell the whole story. A lesser number of sales at a larger price might equal more dollars.

Another factor to consider is the cost of fulfillment. This must be subtracted from total revenue. For example, let’s suppose we have a folio on saving 30% on auto operation. We low it costs us 27? per unit to produce the folio plus 20? to mail it

First Class. So our fulfillment costs are: 47? each. We have tested a 3 price spread: $2.95, $3.95, and $4.95. Results indicate that we can expect a $100 ad investment to yield 100 orders at $2.95 and only 80 orders at $3.95. Results of $100 investment at $4.95 show only 58 orders. So our gross revenue on $100 spent for advertising would be:

As you can see from the above example, $3.95 a unit turned out to be the most advantageous price. But the surprise was that $4.95 worked out netting more dollars than the $2.95 price, even though it didn’t produce as much gross revenue.

Since techniques for testing Direct Mail also apply to your material for “two step” classified conversions, we’ll discuss it first.

Direct Mail lists should always be tested. Nothing could be more disheartening than to have rented a list and then discover that it doesn’t produce enough response to warrant mailing all of it. As explained in the chapter on Direct Mail, you’ll probably

have to rent a minimum of 2,000 or more for your test. This DOES NOT mean that you should’ test the entire amount. I suggest an initial test of 200 to 300, then assess results before testing larger amounts. This allows you to eliminate a “dead” list with- cut having lost two or three hundred dollars mailing to a couple of thousand. If small test results are negative, forget the list. The size of the list has no bearing on the size of your test; however, if you are testing a very small list, the decision as to using the list may be different given the same test results. For example, if your test shows you pulling l l/2% response, which nets you $2.00 per unit sale, then your potential profit is only $30.00 per thousand. On a list of 20,000 it’s certainly not worthwhile to tie up your S for a net of only $600 (20 x 330).

But the decision may well fall the other way if your list is 100,000 or larger.