October 11, 2008

Create control groups to measure your direct mail fulfillment process


Lesson: create a control group for each direct mail campaign to measure the fulfillment process. Use this group to gain insights on the delivery time, physical condition upon receipt, insertion completeness, and more.

Most direct mail campaigns have a 2% success rate. In large businesses, it is common to find direct mail campaigns targeting over 50,000 customers in one mailshot. Email campaigns usually involve a larger number of customers as it’s much cheaper to run.

Mistakes tend to happen. In a recent campaign I was involved in, we sent direct mailers to 4,000 subscribers. It arrived on time, in perfect condition. However, instead of sending a direct mailer containing three inserts, we only sent two. Thankfully, most of the campaign details were delivered. The error was detected by our control group, and was due to an oversight by our fulfillment partner.

With direct mail campaigns, we often overlook the fulfillment process:

  • How long does it take for a direct mailer to reach your customers?
  • What is the condition of the direct mailer upon receipt?
  • Is the quality of the direct mailer good for postal handling?
  • Did you miss out on any insertions?

it’s good practice to create a control group just to measure the effectiveness of your fulfillment process. This also serves as a check and balance if most of the work is outsourced.

In practice:

  • Start with a manageable group – 20 people is sufficient in most cases
  • Find participants in different geographic regions – distributed by city, and even residential type (home / condo)
  • Keep everyone updated on the delivery timeline and ETA
  • Collate feedback with straightforward questions: when did you receive it? creases & folds? did you receive page 1 and 2? etc
  • Use a web-based survey system to collate feedback. Wufoo or SurveyMonkey are worthy candidates for this
  • Reuse this tactic for email campaigns to measure junk mail deliveries, sender and subject line clarity, readability with / without images, etc