Seven years of marketing: one newsletter's track record.

Career Pathways Report, an education newsletter covering school-to-work transitions, is now in its seventh year of publication. In many ways it's a typical business-to-business title, published 26x/year at $177 and targeted to a limited audience of school administrators and other education professionals.

Having been its marketing consultant almost since the beginning, I asked myself, "What have we learned through ten major campaigns and more than 350,000 pieces of mail?"

* Get in the mail. The least successful campaign in CPR's history drew a net profit ratio of $1.10 (more about this one later). CPR publisher Dale Hudelson cofounded Capitol Publications Inc. in the mid-sixties with marketer Ken Callaway who said his greatest contribution to the growth of the company was walking around the office shouting "Now, now, do it now! Don't wait for the last list. Don't rewrite the publisher's letter. Do it now!"

* Conventional wisdom works. With one exception CPR has never varied from the basic newsletter marketing format: #10 window envelope, 4 pp. sales letter, premium buckslip, order form, and reply envelope.

* Seasonality is real. In the education market CPR fall mailings have consistently pulled 1820 percent better than their spring counterparts. Interestingly, they also usually get a "bump" in promotion response rate in late December and early July from prospects who are "managing" their fiscal year budgets.

* Market size. Most b-to-b letters have definable, limited universes. CPR's appears to be somewhere between 25,000 and 35,000 "good" names. Lists added in two campaigns to force-feed the universe to 42,000 and 48,000 names did not work.

* Testing is hard. When the most recent campaign to 25,000 names used 13 lists, you can see that no one list is large enough to yield statistically significant results. A list of 1,700 names can yield 10 orders in one mailing (about $2/$ 1 return) and bomb with two orders the next time (40[cents]). Both responses are within "statistical norms" for that size of mailing.

Hudelson tends to mail any new lists several times before attempting to decide if it works.

* Segmentation. My most ambitious effort. For this campaign we segmented subgroups within the prospect universe and used two different outer envelopes and four custom-tailored first pages to the basic sales letter. A fifth cell received a sample issue mailing test.

This really should have worked. It didn't. Costs for the mailing went up about 25 percent but response didn't budge. It was within 10 percent of the three previous mailings at that time of year.

Why test a sample issue? Because the competitors were all using them. I had the chance to speak with the marketer for one at a newsletter conference. Asked why they used a sample issue, he responded, "Because that's the way Len [Business Publishers Inc. veteran Len Eiserer] does it and he's not going to change."

* Listen to readers. A reader survey brought much higher ratings for how-to stories ("Podunk School Board's 8-step plan for implementing whatsis") than for the shoe-leather intensive, inside-the-beltway ("Meeting in closed mark-up session, the House subcommittee has cut $XX, CPR has learned") stories that editor Hudelson prefers writing. We adjusted both the editorial mix and the sales letter.

* Future. When Hudelson planned the launch in 1993, "school to work" was a trendy idea in education--new federal programs, grant monies, etc. Now, perhaps, it's a more "mature" market. But our last mailing to 25,000 names drew a net profit ratio of $1.82, so we'll still be in there pitching.