How do we know direct mail is actually working?
How can we attribute revenue to postcards, CardalogsTM, or catalogs?
Let’s discuss.
From Postcards to Purchases
The 101: Attributing Direct Mail
At baseline, direct mail attribution comes down to strong data.
You need to be able to tie approximate mail delivery dates with mail recipients’ purchases over a particular attribution window. If you’re able to map this out, you can get a general idea of how what you’ve sent influences buying behavior.
Whether you’re doing this manually or with a mail provider, it’s important to have access to your first-party purchase data via API (e.g., Shopify, SFCC, or BigC) and build a model based on a ~60 day timeframe.
Example of a Dr. Squatch postcard for retention. Note: QR and specific code not shown.
At PostPilot, we provide real-time attribution using these data points.
-
We know approximately when a customer receives a piece of mail.
-
Via our Shopify integration, we know when a customer who has received mail makes a purchase.
-
We typically evaluate campaigns over a 60-day attribution window. If the purchase occurs within 60 days, we attribute it to the mail.
The 201: Maximizing Success
To increase the true impact of this framework, we’re often targeting customers and prospects who are unlikely to engage via email or other channels.
For example, if a customer who has unsubscribed from your email/SMS receives a postcard and purchases 12 days later, it’s a safe assumption that the postcard drove the purchase.
Results from a campaign sent to a group of customers who had unsubscribed from owned channels.
We can be more certain the longer the customer has been unsubscribed and the longer it’s been since they’ve last purchased—due to the lack of exposure to the brand.
But why?
The 301: Why This Works
This is based on classic Recency segmentation (part of broader RF or RFM segmentation), which has been used since the 80s as a powerful predictor of customer behavior.
Recency since last purchase has been shown to be the #1 predictor of subsequent purchases. The longer it’s been since a customer has purchased, the less likely they are to purchase.
If you send mail to a long-unsubscribed customer who hasn’t purchased in between 180–365 days (a segment we commonly recommend), and they purchase within that 60 day attribution window, the most significant brand exposure they’ve had is the mail. They can’t see emails and SMS, and they’re likely not seeing ads.
That lack of exposure similarly applies to acquisition and retargeting campaigns using mail.
Consider a cold prospecting scenario. When we build and send mail to lookalike audiences (i.e., a group of customers similar to a segment of a brand’s current customers), those lookalikes aren’t customers and likely haven’t had any prior exposure to the brand. If the recipient buys something from the brand, we’re certain that the mail drove the purchase.
A Gozney Cardalog used for retargeting anonymous site visitors.
Therefore, we attribute the purchase to it.
Yet we all still yearn for more certainty. Can’t we use coupon codes to track purchases? You can, but they’re not as effective as you’d think.
Problems with Coupon Code Attribution
As marketers, we’d love a “smoking gun” click.
For that reason, brands occasionally ask us about coupon code attribution. There is a perceived increase in attribution certainty.
The truth is that surprisingly few customers actually use the codes on the mail they receive. This makes coupons quite ineffective for tracking the results of your campaign.
The code can be in giant, bright font or take up a substantial portion of the card and still have low use rates.
How do we know?
Our brands have sent tens of thousands of campaigns with countless millions of cards. The data is clear: an extremely high coupon redemption rate is 30%, and it typically falls between 10% and 30%.
That means if you rely on coupons, you’re typically missing at least 70% of your actual conversions. Often, your true ROI is understated by as much as 80%.
Our extensive analysis has shown that coupon ROAS understates true incremental lift by a factor of 3-7x. In other words, the true impact/ROAS of a campaign is 3-7 times the reported coupon ROAS. (And in some cases, it’s significantly higher than 7X.)
For instance, in the above campaign, out of the 6.45% customers who converted, only 22% of them used the discount code that was (unsubtly) shown on the card, despite the fact that the discount was on a premium product.
Granted, when looking at the ROAS attributed to all recipients, it’s possible some of those customers might have purchased anyway even without receiving the card. In certain situations, it can be helpful to run a holdout test to confirm the efficacy.
Direct Your Direct Mail Questions to Us
To learn more about how we think about attribution, or if you'd like to try out full-service direct mail (strategy + design + print + mailing), we’re here for you. We work with brands like HexClad and Dr. Squatch, and we’d love to show you how we can help.
