Attribution is the analysis that credits customer conversions to each of your advertisements. If you come from a digital-first background, you already understand the importance of attribution and a data-driven approach in your campaigns.
In the direct mail world, attribution is measured by comparing:
1. Mailings that you send out with
2. Customer orders that are generated after your mailings have landed in-home
Given both sets of data, our platform will automatically perform that analysis for you in real-time, finding matches where the shipping address, billing address, or email of the orders are the same as those of the mailers you’ve sent out.
We truly believe that direct mail can drive results for your brand - and Poplar enables you to experiment and optimize as you go, with full transparency into the performance of your campaigns. You can execute unlimited multivariate campaigns, test as many offers as you need, take a randomized control group, and review incremental lift in-platform with our attribution reports.
There are a couple of key points to consider when looking at the performance of any direct mail campaign.
It typically takes time for a postcard to land in your customer’s homes, depending on whether you mail First Class or Standard postage. Furthermore, not everyone opens their mail daily, and if they do, will often hold onto marketing material until they are ready to use it. This is why we typically recommend looking at results only starting 30 days after your campaign hits home - and we usually recommend 90 days for DTC and e-commerce brands. In the platform, we let you control your attribution window on a basis of 30, 60, or 90 days.
Oftentimes, people don’t use the offer code! This can be especially true when there are specific terms set around the code - such as the range of products the code is applicable towards, its expiration date, or how compelling the offer is compared to your digital offers. This is why our attribution methodology focuses on data such as the addresses (shipping or billing) and email, which is truly 1:1 and unaffected by the aforementioned other factors.
While attribution for online marketing channels has come a long way, it still struggles with accuracy. Many platforms don’t offer holdout testing, and even those that do can struggle with cross-device connections, cookie blockers, and the never-ending debate over view-through conversions. By measuring attribution using physical addresses, Poplar cuts through any of these struggles to provide accurate 1:1 connections.
In any direct mail campaign, whether it’s your first test or tenth, we highly recommend taking a holdout group. A holdout group is a completely randomized control group taken from the set of people you had intended to mail and could successfully mail.
We recommend taking your holdouts in Poplar, because in Poplar, holdout selection is performed randomly from the group of viable mailings - meaning after address validation and suppressions are taken.
With internal holdouts, the Poplar platform would be able to show you incremental lift metrics for your campaign. These metrics demonstrate the difference in results between your mailed group and holdout group - essentially revealing the true value-add of your campaign.
There are various ways you might look at the effectiveness of any marketing campaign. Below are a few examples of attribution models that can be used.
A first touch attribution model assigns all credit for a conversion to the first advertisement that you sent to a customer. It takes the standpoint that the first advertisement you sent to the customer had the most impact in the conversion. If your main campaign goals are prospecting and customer acquisition, this is a useful model.
A last touch attribution model assigns all credit for a conversion to the last advertisement that you sent to a customer. It takes the standpoint that the latest advertisement you sent to the customer had the most impact in the conversion. After all, it was probably the interaction that resulted in the conversion. This is the model Poplar currently supports.
A linear attribution model assigns credit evenly across all your advertising touchpoints. It takes the standpoint that each advertisement you sent to the customer had an equal impact in the conversion. The caveat, of course, is that the likeliest reality is that not all your channels had an equal influence on the conversion.
A weighted attribution model assigns credit in specified proportions across all your advertising touchpoints.It takes the standpoint that each advertisement you sent to the customer had a specific portion of impact in the conversion.