Read enough articles about digital marketing and you’ll notice something: direct mail often gets casually lumped in with old-school broadcast media. Not interactive, not automated, not testable.
Not true. Not these days. Marketing automation platforms increasingly integrate with direct mail software like Inkit—which means that your direct mail can be just as testable and trackable as any digital campaign.
You can even use A/B testing to improve your direct mail marketing and guide your decision-making. All it takes is a little bit of planning. In this post, I’ll share a few simple direct mail A/B tests that nearly any business can roll out with ease.
But first, let’s get some logistics out of the way.
First Step: Segment Your List
Statistically sound A/B tests require randomness. This is easy on your website, when your A/B testing software can handle that for you.
With direct mail A/B testing, though, you’ll have to create randomized test groups in advance. Here are a few ways to do that.
Route contacts into different automation flows at random
This is the most elegant way to create random A/B test groups, and you generally only need to do it once. The downside: you’ll want to do it well in advance of your first test mailing so that contacts can be sorted into your test buckets as they enter your system.
Neither of Inkit’s marketing automation integration partners currently has an out-of-the-box workflow A/B testing feature, but there are easy-enough workarounds for each platform.
- For HubSpot: Click here to see how to hack any HubSpot form to bucket contacts for A/B testing.
- For Drip: See Step #2 in this post to tag Drip contacts for A/B testing.
How HubSpot thinks about workflow A/B tests
Once you’ve done that, you’ll simply create your two mail templates in Inkit and add the link for each one to the appropriate workflow in HubSpot or Drip.
Don’t use HubSpot or Drip? You’ve still got options.
Export, randomize, segment
If you can export your customer contacts into a CSV file, you can easily use this option. First, copy all the contact data from your spreadsheet and paste it into a randomizer tool—I like this one at Random.org.
Then, copy the reordered data and paste it back into the spreadsheet.
You now have a randomly ordered list of your contacts that you can split neatly in half. Cut and paste half of your spreadsheet’s rows into a new CSV file, and you have two equal segments ready to be uploaded to Inkit (or your mailing solution of choice).
Use Area Mail (plus a little zip code research)
What if you don’t already have customer addresses? You can still run direct mail A/B tests by using location-based targeting, like Inkit’s Area Mail option. Just remember to take the results with an extra dose of skepticism, since your segments won’t be truly random.
Area Mail is sent to all households within zip codes you select. To A/B test an Area Mail campaign, start by identifying pairs of zip codes that are demographically similar to each other (and equally relevant to your business). Assign one to your A group and one to your B group, and continue until you’ve covered enough ground. The average zip code contains about 7,750 households, so you can build segments fairly quickly this way.
Then, add your A-group zip codes to one variation of your Area Mail campaign and your B-group zip codes to the other.
It’s easy to set up an Area Mail campaign inside Inkit.
Once you’ve created your A/B test segments, it’s time to design your direct mail campaign. Here are three test campaign types to consider.
A/B Test #1: Test Two Different Offers
The question: What’s the most effective offer for our customers?
Potential offer types to test
- Dollars-off discount vs. percentage discount
- “Free credit” vs. “free gift certificate”
- Free shipping vs. expedited shipping
- Free trial vs. prorated new-customer discount
- Discount vs. extra perks
How to track results: Create a unique offer code or URL for each test offer and compare conversion rates. Use a statistical significance calculator like this one from Kissmetrics to see how meaningful your results are.
Pro tip: Include an expiration date on your mailing so that you know when to calculate the test results.
A/B Test #2: Test Two Presentations of the Same Offer
The question: What’s the most effective way to frame our offer?
Potential elements to test
- Size formats
- Different emphasis on benefits (such as price vs. convenience, or product quality vs. customer service)
- Different expressions of the same price (such as dollars-off vs. percent saved)
- Different styles of imagery (like the A/B-tested Milwaukee Public Museum postcards below)
How to track results: If you used one of the automated A/B bucketing methods, you don’t need a secondary tracking method: just compare results between the A group and the B group at the end of the test. Otherwise, use one of the methods listed under Test #1.
Pro tip: Don’t just test different postcards—test concepts. You want insights you can apply across your future marketing campaigns, which you won’t get if you change too many elements between variations.
A/B Test #3: Test the Power of Direct Mail
The question: Is it worth adding direct mail to our digital campaigns?
How to test it: You probably know that direct mail can boost the results of your digital campaigns. But will the effect size be worth it for your business?
This style of A/B test gives you an excellent way to answer that question. Here, you’re not testing two different mail pieces—you’re A/B testing two different campaigns, one that includes direct mail and one that doesn’t.
Subscribe both groups to receive the digital elements of your campaign, such as an email sequence. Then, edit one group’s automation workflow to add a postcard (or several) reinforcing the offer.
How to track results: Compare total revenue or activation rates between the groups that did and didn’t receive your mail piece.
Pro tip: Before you run the test, think about the response rate you’re hoping to see for your direct mail campaign. Then, you’ll have an easy benchmark for success when it’s time to add up the results.
We suspect that your results from this type of A/B test will be bright and clear. Still, you should test it for yourself—and prove exactly how well your hard work is paying off.
Got any questions about making direct mail A/B testing work for your business? Ask us in the comments. We’re always happy to help brainstorm solutions!