Direct Mail Packages and A/B Testing

Use this article when

• You want to test different letter versions, carriers, or reply cards against each other.
• You need to split an audience into test and control groups for a direct mail campaign.
• You want to compare response rates across package variants after the drop.

Applies to and prerequisites

How packages and A/B testing work

A package in Direct Mail represents a specific combination of production components: letter version, carrier envelope, reply card, and postage class. Each package gets its own package code for tracking.

An A/B test splits an audience segment so that different portions receive different packages. After the drop, you compare response rates and revenue by package code to identify the winner.

Steps

1. Start a new campaign

1. Go to Direct Mail > Appeals > + Add New Appeal.
2. Complete the Information step: enter Name, Drop date, Appeal code, and select a File layout.
3. Complete the Audience step: add your target segments, exclusions, and seeds.
4. Click Next to reach the Packages step.

2. Create package variants

1. On the Packages step, click Add Package in the top-right of the Packages card.
2. In the package modal, configure the first package:
   • Enter a Description (required)—the human-readable package name shown in Avid and reporting.
   • Enter a Pkg Code (required) for tracking.
   • Set the Letter version (for example, "Emotional appeal").
   • Choose the Carrier envelope type.
   • Select a Reply card style.
   • Pick the Postage class.
   • Enter the Cost per piece.
3. Click Add Package again to create a second variant. Use a different letter version, carrier, or other component so you have a meaningful test.
4. Repeat to add more variants if needed.

3. Enable the A/B split

1. On the package you want to test, turn on Split for A/B Test. An Experiment Treatment Details (Optional) section appears.
2. Configure the treatment details for the variant you are testing.
3. Avid runs a control-vs.-treatment test on that package, so you can compare results by variant after the drop.

4. Assign packages and review

1. Click Next to reach the Selection step.
2. Verify that each audience cell shows the correct package assignment and constituent count.
3. Click Next to reach Review.
4. Confirm totals, package assignments, split percentages, and estimated costs.
5. Click Export to generate the production file.

5. Measure results

After the drop, compare performance across package variants.

1. Allow time for responses to come in—direct mail results typically build over several weeks.
2. Run a Matchback report to attribute gifts to mailing records. See Run a Matchback report.
3. Compare response rate, average gift, and total revenue by package code.
4. Identify the winning variant and use it for your next full-volume mailing.

Result

You created multiple package variants, split your audience for a controlled test, and exported a production file with each recipient assigned to a specific package. After responses arrive, Matchback reporting shows which variant performed best. For reading the results, see Direct Mail Campaign Reporting and Results.

Next steps

• Review results in Insights using the Matchback report. See Run a Matchback report.
• Roll the winning package into your next campaign at full volume.
• Document your findings so future campaigns build on proven results.

Notes and limits

• Keep test groups large enough for statistically meaningful results. Small segments may not produce reliable comparisons.
• Use a distinct Pkg Code for each package so Matchback reporting can attribute results by variant.
• If the audience or packages change after export, re-export the production file so assignments stay accurate.