More Dan Kennedy's Magnetic Marketing Podcast episodes

How to Go Broke with Direct Mail thumbnail

How to Go Broke with Direct Mail

Published 6 Nov 2026

Duration: 42:37

Effective direct mail requires targeted response-based lists, specific formats like oversized postcards, rigorous testing matrices, and long-term customer value analysis to maximize profitability.

Episode Description

You are probably mailing to the wrong people. Dan Kennedy says the quickest way to go broke with direct mail is to mail your entire list. In this podc...

Overview

Craig Simpson, a direct mail strategist who has "mailed out hundreds of millions of sales letters" and whose "largest mailing was 5.7 million pieces of mail," discusses what separates effective direct mail campaigns from those that waste money. He "currently send out about 250 to 300 mail campaigns a year," primarily for acquiring new customers.

Simpson identifies two major list-related mistakes: targeting too broadly, such as running "a list based on men who are age 50 or over," and using compiled lists rather than response-based ones. He recommends running a Merge Purge to remove duplicates and bad addresses, noting he "count on getting rid of about 20% of those names in the Merge Purge" for smaller mailings. On format, he favors short pieces like the Oversize Postcard and four-page self-mailer for driving prospects online, while longer formats like the digest (up to "88 pages") work for direct sales. For testing, he advises against 50-50 splits, suggesting instead mailing "70,000 pieces the control and the remaining 20,000 pieces the test."

A recurring theme is the importance of long-term customer value. Simpson argues the "biggest mistakes made with direct mail" involve companies looking only at first-transaction economics rather than lifetime value, noting that in one campaign "each of those customers went on to spend over $1,000." He also stresses setting up tracking in advance so marketers can identify which lists produce results and which should be discarded.

What If

  • What if you mailed a four-page self-mailer to a response-based list of 3,000 names in your exact software niche instead of running broad digital ads?

    • Move: Rent a response-based list (not a compiled one) of 3,000 names, design a four-page self-mailer - "a sheet of 11 by 17 paper folded in half twice" - driving recipients to your online demo.
    • Why Now?: The source stresses "who you mail can matter more than what you mail" and warns that compiled lists contain people who "have not raised their hand to a prior direct mail campaign." A solo dev can afford a small test batch.
    • Expected Upside (proposal): You could identify whether physical direct response converts in your niche before committing to larger spend, mirroring the source's advice to "only take 3,000 of those names and test it."
  • What if you allocated the bulk of your test mailing to your proven control and kept the experimental variable small, as the source recommends "70,000 pieces the control and the remaining 20,000 pieces the test"?

    • Move: Run both your current best-performing channel and a new list or format simultaneously, weighting the majority toward the control and tracking response rates separately.
    • Why Now?: The source cautions that a 50-50 split risks that "you've just wasted half of your mailing" if the test bombs, and recommends the 70,000/20,000 split to limit downside.
    • Expected Upside (possibility): You preserve most of your revenue while still gathering "response percentage" data on the new variable without exposing your entire budget to an unproven channel.
  • What if you set your customer-acquisition cost threshold from six-month lifetime value rather than first-purchase margin?

    • Move: Calculate what a typical subscriber spends over six months, then set your per-customer acquisition budget at roughly half that figure before choosing channels.
    • Why Now?: The source describes a case where "We were selling a course for $195, and at that time, I was willing to pay $300 to $350 to acquire a customer" yet "each of those customers went on to spend over $1,000."
    • Expected Upside (proposal): For recurring-revenue software, you could justify higher per-unit costs and "be more aggressive" in channels you'd otherwise dismiss as unprofitable on first sale alone.

Takeaway

  • Segment your prospect list by specific behavioral and demographic traits (income, region, prior purchases) rather than broad categories; the source warns that "men who are age 50 or over" misses "a lot of other characteristics about those men."
  • Source response-based lists - people who've already acted on a prior campaign - instead of compiled demographic lists where "the people on the list haven't responded to anything, they have not raised their hand to a prior direct mail campaign."
  • Run a Merge Purge before every mailing; expect to remove about 20% of names on smaller mailings and "30 or 40%" on larger ones due to duplicates and bad addresses.
  • Set up tracking before launching any campaign so you can identify which lists are "home runs" versus "dogs" and reallocate spend accordingly.
  • Calculate customer lifetime value before setting your acquisition cost threshold; if LTV is "$400," you can "spend $200 to acquire a customer" even when the first transaction shows a loss.

Recent Episodes of Dan Kennedy's Magnetic Marketing Podcast

23 Oct 2026 Balancing Brand and Direct Response Advertising

"Challenges B2B vs. B2C marketing distinctions, advocating for niche-specific strategies that recognize similar buyer behaviors, decision-making complexities, and the need for high-level executive targeting."

9 Oct 2026 How to Make Yourself Fascinating So People Seek You Out

"Explores how fascination drives engagement and persuasion through seven triggers - power, mystique, trust, lust, alarm, vice, and prestige - showing how businesses can leverage these to captivate audiences, justify premium pricing, and build stronger connections."

More Dan Kennedy's Magnetic Marketing Podcast episodes