Email Targeting's Impact on ROI
Sometimes the best return on investment strategy yields a counterintuitive outcome. I once worked with an apparel company that offered men’s and women’s merchandise to customers via email marketing campaigns. This company had an algorithm that weighted prior purchases and customer preferences, using both pieces of information to determine which version of an email campaign the customer would receive.
The algorithm used results from mail and holdout groups to optimize results. This is where things get interesting: A segment of customers who spent 70 percent of their historical demand on men’s merchandise and who asked the company to receive men’s email marketing messages was placed into one of three test groups. The first test group received an email featuring men’s merchandise. The second group received an email featuring women’s merchandise. The third group didn't receive an email. Here are the results (by average ROI/email):
- men’s marketing message = $4.75
- women’s marketing message = $4.95
- no email = $4.55
Here’s a customer who spends a majority of money on men’s merchandise and specifically requests to receive men’s email marketing messages, yet generates more revenue when targeted with a women’s email marketing message.
What would you, the chief marketing officer, recommend as an appropriate marketing strategy for this customer going forward?
There are two strategic questions you have to ask yourself when trying to maximize ROI:
- Are you willing to honor a customer request even if it means that your company makes less profit?
- Are you willing to allow computer algorithms and optimization testing drive your business to a more profitable strategy that, on the surface, doesn’t make intuitive sense?
I’ve always struggled with the answers to each question. If the customer asks to receive men’s email marketing messages, then I need to send the customer men’s email marketing messages — and I need to quantify the amount of profit I’m going to lose because I choose to honor customer preferences. I wish I could say I’ve always honored customer preferences.
The second question is far more strategic. Do you trust test results that yield counterintuitive outcomes? Do you send a men’s customer a women’s email marketing message because your test results suggest it'll be more profitable to do so?
I tend to side with what testing and optimization suggest is the optimal strategy. That being said, this is one place where marketing tactics earn a seat at the executive table. Business leaders need to know about situations where customers act in counterintuitive manners and understand how much profit is at risk if the business doesn't capitalize on counterintuitive behaviors.